<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:itunes="http://www.itunes.com/dtds/podcast-1.0.dtd" xmlns:googleplay="http://www.google.com/schemas/play-podcasts/1.0"><channel><title><![CDATA[Informerad.se]]></title><description><![CDATA[Writing about how data, AI, and digital strategy actually work in real organizations — beyond the hype, inside the execution.]]></description><link>https://www.informerad.se</link><image><url>https://substackcdn.com/image/fetch/$s_!tzCa!,w_256,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd0ffb9fc-0ec2-4a2a-9613-7f35a46467e9_1024x1024.png</url><title>Informerad.se</title><link>https://www.informerad.se</link></image><generator>Substack</generator><lastBuildDate>Mon, 18 May 2026 03:56:25 GMT</lastBuildDate><atom:link href="https://www.informerad.se/feed" rel="self" type="application/rss+xml"/><copyright><![CDATA[Johan Åhlén]]></copyright><language><![CDATA[en]]></language><webMaster><![CDATA[informerad@substack.com]]></webMaster><itunes:owner><itunes:email><![CDATA[informerad@substack.com]]></itunes:email><itunes:name><![CDATA[Johan Åhlén]]></itunes:name></itunes:owner><itunes:author><![CDATA[Johan Åhlén]]></itunes:author><googleplay:owner><![CDATA[informerad@substack.com]]></googleplay:owner><googleplay:email><![CDATA[informerad@substack.com]]></googleplay:email><googleplay:author><![CDATA[Johan Åhlén]]></googleplay:author><itunes:block><![CDATA[Yes]]></itunes:block><item><title><![CDATA[Winfried Adalbert Etzel: The governance playbook is broken — here's what to do instead]]></title><description><![CDATA[The author of Data Governance in the Wild on why classical governance is already broken, why accountability beats control &#8212; and why treating AI governance as a separate discipline is itself a governan]]></description><link>https://www.informerad.se/p/winfried-adalbert-etzel-the-governance</link><guid isPermaLink="false">https://www.informerad.se/p/winfried-adalbert-etzel-the-governance</guid><dc:creator><![CDATA[Johan Åhlén]]></dc:creator><pubDate>Wed, 25 Mar 2026 09:44:38 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!H7JU!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ef93ff1-837d-4204-a35b-2ad82dba13d8_1080x949.jpeg" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!H7JU!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ef93ff1-837d-4204-a35b-2ad82dba13d8_1080x949.jpeg" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!H7JU!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ef93ff1-837d-4204-a35b-2ad82dba13d8_1080x949.jpeg 424w, https://substackcdn.com/image/fetch/$s_!H7JU!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ef93ff1-837d-4204-a35b-2ad82dba13d8_1080x949.jpeg 848w, https://substackcdn.com/image/fetch/$s_!H7JU!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ef93ff1-837d-4204-a35b-2ad82dba13d8_1080x949.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!H7JU!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ef93ff1-837d-4204-a35b-2ad82dba13d8_1080x949.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!H7JU!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ef93ff1-837d-4204-a35b-2ad82dba13d8_1080x949.jpeg" width="1080" height="949" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/3ef93ff1-837d-4204-a35b-2ad82dba13d8_1080x949.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:949,&quot;width&quot;:1080,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:223389,&quot;alt&quot;:&quot;Winfried Adalbert Etzel&quot;,&quot;title&quot;:null,&quot;type&quot;:&quot;image/jpeg&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://www.informerad.se/i/192069169?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc7111348-2e5b-465d-8b60-67edf67b6719_1080x1080.jpeg&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="Winfried Adalbert Etzel" title="Winfried Adalbert Etzel" srcset="https://substackcdn.com/image/fetch/$s_!H7JU!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ef93ff1-837d-4204-a35b-2ad82dba13d8_1080x949.jpeg 424w, https://substackcdn.com/image/fetch/$s_!H7JU!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ef93ff1-837d-4204-a35b-2ad82dba13d8_1080x949.jpeg 848w, https://substackcdn.com/image/fetch/$s_!H7JU!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ef93ff1-837d-4204-a35b-2ad82dba13d8_1080x949.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!H7JU!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ef93ff1-837d-4204-a35b-2ad82dba13d8_1080x949.jpeg 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image"><svg role="img" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg"><g><title></title><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a><figcaption class="image-caption"><em>Winfried Adalbert Etzel</em></figcaption></figure></div><p><em>Winfried Adalbert Etzel is a data governance professional and author based in Stavanger, Norway. With his background in history, political science, and law &#8212; and years working across Norwegian public administration and the oil and gas industry &#8212; he brings an unusual intellectual lens to a discipline most organisation still get fundamentally wrong. His forthcoming book,</em> Data Governance in the Wild, <em>argues for a fundamental rethinking of how organisations govern their data in an era of distributed landscapes and AI.</em></p><p><em>We spoke about what governance actually means, why the Theranos scandal is a data governance story, and why the assumptions underpinning most governance programmes have already quietly collapsed.</em></p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.informerad.se/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Informerad.se! Subscribe for free to receive new posts and support my work.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><h3>From shoe store to archives to data governance</h3><p>The journey to becoming one of the Nordic&#8217;s more distinct voices in data governance began, of all places, in a shoe store in Oslo.</p><p>After studying history, political science, and law at universities in Germany &amp; Norway, Winfried Adalbert Etzel spent a few years managing a shoe store. There he saw how data affects the business. He then joined a consultancy working with the National Archives of Norway &#8212; evaluating what public institutions needed to preserve and how to prepare it for archival.</p><p>&#8220;I always felt like I was working on the wrong end of the lifecycle,&#8221; he says. &#8220;I was just dealing with issues that could have been avoided if better governance had been in place from the beginning.&#8221; That frustration pushed him earlier and earlier into the data lifecycle. Then he met his wife and moved to Stavanger. There he moved into consulting in the oil and gas industry and went deeper and deeper into data governance.</p><p>The academic background never left him. &#8220;If you study history,&#8221; he says, &#8220;you learn a lot about source criticism &#8212; how to evaluate a text written by someone, for a certain purpose, at a certain point in time.&#8221; He pauses. &#8220;We&#8217;re doing exactly the same with data and AI. Why was that model trained? On what dataset? What was the context? It&#8217;s forensics.&#8221;</p><p>Political science, too, shapes his thinking &#8212; particularly when he&#8217;s designing federated governance models. &#8220;I always use the European Union as an example,&#8221; he says. &#8220;How you set up structures, how responsibilities are shared &#8212; there&#8217;s so much good material in political science for that.&#8221;</p><h3>What governance is actually for</h3><p>The central provocation of Etzel&#8217;s forthcoming book is direct: the classical model of data governance &#8212; built around authority, control, and centralised oversight &#8212; no longer works. &#8220;It&#8217;s outdated. We need new mechanisms, a new way of thinking.&#8221; Most reform-minded writing in the field, he argues, quickly slides into execution: tooling, monitoring, catalogues, policies. &#8220;That leads you away from the fundamental question &#8212; why are we doing this in the first place?&#8221;</p><p>His answer is distilled into a definition he spent considerable time crafting, deliberately aligned with how ISO standards define corporate governance: data governance is a <em>human-based system by which data assets in a socio-technical system are directed, overseen, and by which the organization is held accountable for achieving its defined purpose.</em></p><p>From this definition he draws what he calls the <em>data governance triad</em>: three enduring roles that apply whether an organisation has sixty employees or sixty thousand. The first is the <em>data negotiator</em> &#8212; someone who understands and translates between business needs, regulatory demands, technology constraints and market realities. The second is the <em>data director</em> &#8212; focused not on tomorrow&#8217;s to-do list, but on where the organisation wants to be in five years&#8217; time. The third is the <em>data auditor</em> &#8212; the one who ensures accountability sticks.</p><p>&#8220;No matter how complicated your technology stack is,&#8221; he says, &#8220;you always need to do these three things. They&#8217;re enduring.&#8221;</p><h3>The accountability gap &#8212; and what Theranos tells us</h3><p>Of the three roles in the triad, accountability is the one Etzel returns to most forcefully. It is also, he argues, the most consistently undervalued.</p><p>&#8220;We talk a lot about data ownership,&#8221; he says. &#8220;Roles and responsibilities, who owns the data. That&#8217;s all well and nice. But who is ultimately accountable for any outcomes &#8212; whether produced by a machine or a human?&#8221;</p><p>He reaches for an example most people will recognise: Theranos, the blood-testing company whose founder Elizabeth Holmes raised billions on technology that never existed. &#8220;The data showed constantly that it didn&#8217;t work. People who spoke up were silenced. The board chose to believe the CEO rather than the actual data.&#8221;</p><div class="pullquote"><p>&#8220;The board chose to believe the CEO rather than the actual data.&#8221;</p></div><p>For Etzel, the Theranos collapse is not primarily a story about fraud or charisma &#8212; it&#8217;s a governance story. &#8220;If you had placed accountability clearly in the organisation, you could have avoided all of that. <em>The board didn&#8217;t take their accountability.</em> That&#8217;s what let it run for so long.&#8221;</p><h3>AI governance is not a separate discipline</h3><p>Where most organisations are busy launching AI policies, AI oversight boards, and AI ethics frameworks &#8212; Etzel makes an argument that cuts against the grain: AI governance isn&#8217;t a new discipline. It&#8217;s an acceleration of something that has been developing for over a decade.</p><p>&#8220;Before 2022, before LLMs and generative AI made the headlines, we were already moving towards distributed landscapes, federated data governance, computational governance,&#8221; he says. &#8220;AI came around and everyone said, &#8216;now we need AI governance.&#8217; But it&#8217;s just that same development, moving faster.&#8221;</p><p>He goes further still. Rather than treating data governance, AI governance, IT governance, risk governance and corporate governance as separate disciplines that need to be &#8220;aligned,&#8221; he argues they should be understood as facets of a single thing &#8212; and governed as one, from the top.</p><p>&#8220;If you treat them as separate disciplines, you create gaps. And those gaps have been issues for data governance for years. Now with AI, we&#8217;re just creating another blind spot.&#8221;</p><h3>The federated dilemma</h3><p>The book grapples at length with a tension that has frustrated practitioners for years: <em>federated, domain-driven governance</em> on one hand versus <em>centralised control</em>.</p><p>The data mesh movement, he notes, borrowed heavily from domain-driven design in software engineering. The problem is that data doesn&#8217;t behave like code. &#8220;Data is reused, reinterpreted, recombined across domains. It doesn&#8217;t respect bounded contexts in the same way.&#8221; The result is predictable: a change in one corner of the organisation triggers ripple effects everywhere, ownership disputes erupt, and a central coordination team ends up fighting battles on every front simultaneously.</p><p>But the alternative &#8212; insisting on uniform definitions and centralised control across a large organisation &#8212; is equally unrealistic. &#8220;In large organisations, different cultures, different teams, different domains &#8212; you&#8217;re never going to get everyone aligned on the same customer definition.&#8221;</p><p>The answer, as Etzel frames it, is learning to balance autonomy and alignment rather than choosing between them. &#8220;You need to respect autonomy and create alignment where it&#8217;s genuinely needed &#8212; not central adherence everywhere, and not chaos.&#8221;</p><h3>Data as asset &#8212; not a service desk</h3><p>One of Etzel&#8217;s sharper critiques is directed at organisations that have drifted into thinking about data in purely service-function terms &#8212; a legacy, he argues, of how IT departments were originally structured.</p><p>&#8220;In the nineties, IT had enormous power. If you wanted anything changed, you punched a ticket. When data emerged as its own discipline, we just copied that mindset.&#8221; The result: data teams became sophisticated help desks, handling requests for dashboards and nightly jobs.</p><p>The alternative frame &#8212; data as an asset &#8212; has been circulating since at least The Economist&#8217;s famous 2017 comparison of data to oil. But Etzel thinks it has never fully landed, partly because it remains abstract. He prefers a financial analogy: &#8220;Do you know what your organisation uses its money for? Any controller would know that &#8212; tracking financial assets is literally their job. With data, we never got there. We never built the equivalent function.&#8221;</p><p>Data products, in his view, are the bridge. &#8220;A data product is a way of packaging the asset to make it consumable, understandable &#8212; with structure, validation, quality, accessibility. Things you wouldn&#8217;t get from an abstract concept.&#8221;</p><h3>How to start &#8212; and what to stop doing</h3><p>For organisations beginning their governance journey now, Etzel is direct: stop reaching for blueprints.</p><p>&#8220;A lot of people try to sell blueprint data governance. It doesn&#8217;t exist. It&#8217;s entirely dependent on how your organisation is set up, who your customers are, how many people are working there.&#8221; His recommended first move is simply to orient yourself &#8212; understand how the organisation works, the pain points, and what it actually uses data for.</p><div class="pullquote"><p>&#8220;A lot of people try to sell blueprint data governance. It doesn&#8217;t exist.&#8221;</p></div><p>The second step is to surface a single problem: a contested definition, a quality issue, an ownership dispute. &#8220;Don&#8217;t try to boil the ocean. Solve one thing. Produce something tangible &#8212; a policy, a data contract, a catalogued dataset. Then feed that back into the organisation and ask: could this approach work for the next problem?&#8221;</p><p>That cycle &#8212; <em>orient, surface, produce, feedback</em> &#8212; is, in his view, both more durable and more persuasive than trying to launch a comprehensive governance programme from the start.</p><h3>Three broken assumptions</h3><p>His most consequential argument isn&#8217;t about tools or frameworks. It&#8217;s about the ground shifting beneath them.</p><p>The first broken assumption is stability. &#8220;We enact our governance as if the environment is stable and will stay that way for ten or fifteen years. It&#8217;s not. It&#8217;s fragile.&#8221;</p><p>The second is controllability &#8212; the belief that a central governance function can actually oversee how data flows and is defined across a large organisation. &#8220;You can&#8217;t. You are entirely reliant on the people actually working with the data.&#8221;</p><p>The third assumption is the one he finds most significant: proximity. The idea that people are close enough to their data to understand it, to notice when something is wrong, to take meaningful ownership of it. &#8220;I think that&#8217;s been gone for years. We&#8217;ve complicated ourselves away from it &#8212; and AI is accelerating that further.&#8221;</p><p>What Etzel hopes to see instead: governance frameworks built around resilience and adaptability rather than stability and control, with AI being used in the governance mechanisms themselves &#8212; not just as the thing being governed.</p><p>&#8220;AI can make our work much easier,&#8221; he says. &#8220;Or much more complicated. But it doesn&#8217;t change why governance exists in the first place. That&#8217;s important to remember.&#8221;</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.informerad.se/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Informerad.se! Subscribe for free to receive new posts and support my work.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><p><em>Note! The content on this blog reflects my personal opinions and does not represent my employer. As the publisher, I am not responsible for the comments section. Each commenter is responsible for their own posts.</em></p>]]></content:encoded></item><item><title><![CDATA[Carl-Johan Nakamura: Why enterprise AI keeps failing — and what leaders can do about it]]></title><description><![CDATA[Serial entrepreneur and former Chief Data Officer Carl-Johan Nakamura on why enterprise AI keeps failing, the brutal first weeks of a CDO - and why regulation may finally force boards to pay attention]]></description><link>https://www.informerad.se/p/carl-johan-nakamura-why-enterprise</link><guid isPermaLink="false">https://www.informerad.se/p/carl-johan-nakamura-why-enterprise</guid><dc:creator><![CDATA[Johan Åhlén]]></dc:creator><pubDate>Mon, 09 Mar 2026 06:52:16 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!QR6V!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Faf8d89e0-efd5-4360-afab-bd179a835008_3091x2073.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!QR6V!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Faf8d89e0-efd5-4360-afab-bd179a835008_3091x2073.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!QR6V!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Faf8d89e0-efd5-4360-afab-bd179a835008_3091x2073.png 424w, https://substackcdn.com/image/fetch/$s_!QR6V!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Faf8d89e0-efd5-4360-afab-bd179a835008_3091x2073.png 848w, https://substackcdn.com/image/fetch/$s_!QR6V!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Faf8d89e0-efd5-4360-afab-bd179a835008_3091x2073.png 1272w, https://substackcdn.com/image/fetch/$s_!QR6V!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Faf8d89e0-efd5-4360-afab-bd179a835008_3091x2073.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!QR6V!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Faf8d89e0-efd5-4360-afab-bd179a835008_3091x2073.png" width="1456" height="976" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/af8d89e0-efd5-4360-afab-bd179a835008_3091x2073.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:976,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:5302086,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://www.informerad.se/i/189766889?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Faf8d89e0-efd5-4360-afab-bd179a835008_3091x2073.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!QR6V!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Faf8d89e0-efd5-4360-afab-bd179a835008_3091x2073.png 424w, https://substackcdn.com/image/fetch/$s_!QR6V!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Faf8d89e0-efd5-4360-afab-bd179a835008_3091x2073.png 848w, https://substackcdn.com/image/fetch/$s_!QR6V!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Faf8d89e0-efd5-4360-afab-bd179a835008_3091x2073.png 1272w, https://substackcdn.com/image/fetch/$s_!QR6V!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Faf8d89e0-efd5-4360-afab-bd179a835008_3091x2073.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image"><svg role="img" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg"><g><title></title><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p><em>Carl-Johan Nakamura has spent nearly two decades at the intersection of data, AI, and enterprise leadership &#8212; serving as CIO, COO, and CDO at organizations including Siemens Healthcare, IBM, and ZEISS. Today he is CEO of <a href="https://www.ai81works.com/">AI81 Works</a>, an executive search and consultancy firm helping companies find the right talent to make AI work in practice.</em></p><p><em>We spoke about what&#8217;s really holding enterprise AI back, and what it will take to change it &#8212; from boardroom culture to the first three weeks on the job as a CDO.</em></p><h3><strong>From data mining to the first unofficial CDO office</strong></h3><p>Nakamura&#8217;s path into data and AI began in 2007, when he was managing software implementations at US hospitals for Siemens Healthcare. &#8220;Back then it wasn&#8217;t called AI or machine learning &#8212; it was called data mining and decision support systems,&#8221; he recalls. What struck him early was a lesson that remains central to his thinking today: &#8220;if the data foundation isn&#8217;t right, nothing built on top of it can be trusted.&#8221;</p><p>One experience in particular became a turning point. Three different teams within Siemens were measuring the same customer satisfaction KPI against the same dataset &#8212; and arriving at three different ratings: red, yellow, and green. &#8220;That inconsistency became my entry point into business intelligence and analytics,&#8221; he says.</p><p>What followed was unusual for someone inside a large corporation. Nakamura wrote a 45-page business plan &#8212; the kind of thing you&#8217;d expect from a startup founder &#8212; for what he called the &#8220;Office of the Chief Data Officer.&#8221; He went door to door pitching it to senior executives, and then it became the first unofficial CDO office at Siemens Health Services.</p><h3><strong>Why &#8216;81%&#8217; &#8212; and the cost of getting the hire wrong</strong></h3><p>The name of Nakamura&#8217;s company encodes a core conviction: for AI to truly work in any organization, at least 81% of the people need to be engaged, educated, and genuinely living its value. &#8220;You need 81% of the people drinking the Kool-Aid,&#8221; he says &#8212; half-jokingly, but with a serious point. That belief became the name AI81 Works.</p><p>The company operates primarily as an executive search and strategy consultancy firm, helping organizations fill data and AI leadership roles. Nakamura is blunt about why this matters: the field has enormous churn, and most companies underestimate the cost. A data governance leader who joins for two and a half years typically needs <strong>six months to onboard</strong> and <strong>another six to learn the local context</strong>. Add the reality that many decide to leave <strong>months before they actually exit</strong>, and the company may be left with barely <strong>12&#8211;15 months</strong> of truly productive value. Getting the hire right the first time isn&#8217;t optional &#8212; it&#8217;s existential.</p><h3><strong>The iceberg problem &#8212; and the data janitor trap</strong></h3><p>When asked where the biggest friction points lie, Nakamura returns to what he calls the iceberg analogy. Executive boards across industries are investing heavily in the glamorous, visible tip &#8212; the AI models, the chatbots, the dashboards. Meanwhile, the vast data foundation underneath goes ignored.</p><p>&#8220;The perception building up among executives is that AI can solve pretty much anything.&#8221; Technically, there is enormous potential. But applying these tools inside enterprises burdened with decades of process debt and data debt &#8212; organizational, not just technical &#8212; is a different story. &#8220;It&#8217;s no wonder that <strong>figures as high as 95%</strong> get cited for enterprise AI project failure,&#8221; he says. &#8220;That&#8217;s with emphasis on enterprise AI.&#8221;</p><p>He also describes what he calls the &#8220;data janitor trap&#8221; &#8212; a pattern he has encountered in every CDO role, whether leading a team of 10 or 500. Colleagues would approach him with requests better suited to a help desk: a nightly job that didn&#8217;t run, a dashboard that won&#8217;t refresh. &#8220;The CDO role gets reduced to being the Chief Dashboard Officer &#8212; when it should be about strategic enablement.&#8221;</p><h3><strong>The first three weeks define everything</strong></h3><p>Nakamura believes the well-known Gartner advice about &#8220;the first 100 days&#8221; is far too generous. &#8220;As a first-time or second-time CDO in a legacy business model and company, your success or failure will be evaluated in the first three weeks,&#8221; he says.</p><p>He speaks from experience. When he was appointed as the first CDO at ZEISS, his approach was to spend his first weeks listening rather than pitching &#8212; meeting the CEO, the CFO who had hired him, and key business-side SVPs, letting them do most of the talking.</p><p>The strategy paid off. By the end of week one, he had built relationships with several influential leaders, earning an unexpected invitation to address the executive board in just his second week. Those early moves positioned him as a strategic enabler rather than a data janitor stuck on defense.</p><h3><strong>The executive AI literacy gap</strong></h3><p>Nakamura sees a broader transformation underway in what it means to be a business executive. &#8220;The modern business executive won&#8217;t be successful doing what they used to do,&#8221; Nakamura says. &#8220;The modern executive needs to dramatically raise their AI literacy.&#8221;</p><p>Yet almost nobody at the top is doing so. He recently participated in discussions about bringing a top-tier university&#8217;s corporate board AI program to Europe, and the pattern was clear: mid-level leaders attend, but at the board level, almost nobody signs up. &#8220;<strong>They think it doesn&#8217;t touch them</strong>,&#8221; he says.</p><p>That mindset is starting to crack &#8212; driven primarily by regulation. The EU AI Act, Canada&#8217;s AIDA, and the US NIST framework are forcing boards to engage with AI governance. &#8220;Europe is taking this more seriously from a responsible and ethical perspective than anyone else,&#8221; he says. &#8220;Just as we saw with GDPR but this time with real teeth, he hopes.&#8221;</p><h3><strong>Two bets for the next 18 months</strong></h3><p>Looking ahead, Nakamura identifies two factors that will separate leaders from laggards.</p><p>The first is responsible AI as a brand. As personal data grows more valuable and data ownership more contested, organizations that embed responsible AI into their corporate DNA &#8212; and communicate it externally &#8212; will gain both trust and competitive edge.</p><p>The second is agentic fluency. As agentic AI matures, winning organizations will move beyond &#8220;data literacy&#8221; toward something more fundamental: orchestrating human and AI workforces together. &#8220;Some firms are already hiring agentic AI systems, giving them employee ID numbers, and onboarding them alongside human employees,&#8221; he notes.</p><p>But none of this happens by default. In Nakamura&#8217;s view, it takes senior leaders who <strong>educate themselves first</strong> &#8212; and then drive the capability through the organization.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.informerad.se/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Informerad.se! Subscribe for free to receive new posts and support my work.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><p><em>Note! The content on this blog reflects my personal opinions and does not represent my employer. As the publisher, I am not responsible for the comments section. Each commenter is responsible for their own posts.</em></p>]]></content:encoded></item><item><title><![CDATA[Stop running AI as projects—start running AI as an investment portfolio]]></title><description><![CDATA[If individual AI initiatives don&#8217;t meet expectations, it&#8217;s all too easy to conclude that &#8220;AI doesn&#8217;t work for us.&#8221; But most of the time, the problem is the governance model.]]></description><link>https://www.informerad.se/p/stop-running-ai-as-projectsstart</link><guid isPermaLink="false">https://www.informerad.se/p/stop-running-ai-as-projectsstart</guid><dc:creator><![CDATA[Johan Åhlén]]></dc:creator><pubDate>Wed, 25 Feb 2026 07:07:19 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!IUP9!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F38cdb044-e968-4e71-ab9f-56b948ffe965_2752x1536.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p>Many organizations start with AI like this: pick a promising use case, write a business case, present it to leadership, run a pilot, realize the expectations weren&#8217;t met&#8212;and then question the entire AI effort.</p><p>But AI doesn&#8217;t behave like &#8220;regular&#8221; projects. Outcomes for individual bets are often uncertain and the value is unevenly distributed. On top of that, most of the value sits in what happens around the technology: changed ways of working, better decisions, and new business opportunities.</p><p>You can see it in the numbers:</p><ul><li><p><strong>PwC&#8217;s 2026 Global CEO Survey:</strong> <a href="https://www.pwc.com/gx/en/news-room/press-releases/2026/pwc-2026-global-ceo-survey.html">only </a><strong><a href="https://www.pwc.com/gx/en/news-room/press-releases/2026/pwc-2026-global-ceo-survey.html">12%</a></strong><a href="https://www.pwc.com/gx/en/news-room/press-releases/2026/pwc-2026-global-ceo-survey.html"> of CEOs</a> say AI has so far delivered both cost and revenue benefits, and <strong>56%</strong> still see no significant financial benefit.</p></li><li><p><strong>BCG&#8217;s 2025 study</strong> (over <strong>1,250</strong> companies): <a href="https://media-publications.bcg.com/The-Widening-AI-Value-Gap-Sept-2025.pdf">only </a><strong><a href="https://media-publications.bcg.com/The-Widening-AI-Value-Gap-Sept-2025.pdf">5%</a></strong> create AI value &#8220;at scale,&#8221; while <strong>60%</strong> see no material value at all despite investments.</p></li><li><p><strong>IDC/Lenovo study</strong> (cited by CIO): <strong><a href="https://www.cio.com/article/3850763/88-of-ai-pilots-fail-to-reach-production-but-thats-not-all-on-it.html">88%</a></strong><a href="https://www.cio.com/article/3850763/88-of-ai-pilots-fail-to-reach-production-but-thats-not-all-on-it.html"> of AI pilots</a> don&#8217;t reach broad deployment. For every <strong>33</strong> pilots, only <strong>4</strong> make the jump to production.</p></li></ul><p>If you run AI like traditional projects, the result is often a pilot graveyard. The first pilots become a &#8220;trial&#8221;&#8212;and if they don&#8217;t scale, the trial also becomes the verdict: &#8220;The AI initiative was a bad idea.&#8221;</p><p>AI advisor and author <strong>Tobias Zwingmann</strong> captures exactly this pattern in his article <em><a href="https://blog.tobiaszwingmann.com/p/ai-roadmaps-beat-ai-projects">Why AI Roadmaps beat AI projects</a></em>. He describes how getting the first budget usually isn&#8217;t the hard part&#8212;but the initiative can die when the first prototype doesn&#8217;t deliver what people hoped. His reasoning aligns closely with my own experience and is an important starting point for this article.</p><p>The solution is to change the governance model: AI needs to be run like an investment portfolio, with multiple initiatives, shared foundational capabilities, and a roadmap where shut-down initiatives become lessons rather than failures.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!IUP9!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F38cdb044-e968-4e71-ab9f-56b948ffe965_2752x1536.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!IUP9!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F38cdb044-e968-4e71-ab9f-56b948ffe965_2752x1536.png 424w, https://substackcdn.com/image/fetch/$s_!IUP9!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F38cdb044-e968-4e71-ab9f-56b948ffe965_2752x1536.png 848w, https://substackcdn.com/image/fetch/$s_!IUP9!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F38cdb044-e968-4e71-ab9f-56b948ffe965_2752x1536.png 1272w, https://substackcdn.com/image/fetch/$s_!IUP9!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F38cdb044-e968-4e71-ab9f-56b948ffe965_2752x1536.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!IUP9!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F38cdb044-e968-4e71-ab9f-56b948ffe965_2752x1536.png" width="1456" height="813" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/38cdb044-e968-4e71-ab9f-56b948ffe965_2752x1536.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:813,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:5850025,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://www.informerad.se/i/189107391?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F38cdb044-e968-4e71-ab9f-56b948ffe965_2752x1536.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!IUP9!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F38cdb044-e968-4e71-ab9f-56b948ffe965_2752x1536.png 424w, https://substackcdn.com/image/fetch/$s_!IUP9!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F38cdb044-e968-4e71-ab9f-56b948ffe965_2752x1536.png 848w, https://substackcdn.com/image/fetch/$s_!IUP9!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F38cdb044-e968-4e71-ab9f-56b948ffe965_2752x1536.png 1272w, https://substackcdn.com/image/fetch/$s_!IUP9!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F38cdb044-e968-4e71-ab9f-56b948ffe965_2752x1536.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image"><svg role="img" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg"><g><title></title><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><h3>When the first pilot becomes the verdict</h3><p>The pattern is often predictable.</p><p>An idea shows potential. Many people see the possibilities. Then reality hits: data access, integration, information security, legal constraints, operations/governance, training, and process change.</p><p>When this gets difficult, it&#8217;s easy to mistake friction for final proof. That&#8217;s when having the right governance model becomes critical.</p><p>If leadership starts asking, &#8220;Should we really continue?&#8221; the initiative has often already lost momentum. Not because AI lacks potential, but because you built a governance model where too much depended on the first attempt.</p><p>Eventually, the organization might buy Copilot licenses and feel like that&#8217;s &#8220;enough&#8221; AI.</p><h3>Why AI differs from IT</h3><p>In much of IT (<a href="https://www.informerad.se/p/utility-it-and-strategic-it-two-very">utility IT</a>), you get impact by implementing a technical solution and rolling it out. It&#8217;s not always easy&#8212;but it&#8217;s often fairly planable.</p><p>AI, on the other hand, works more like business transformation. Value emerges when everyday behaviors change, when decisions improve, and when workflows are redesigned.</p><p>That means entirely different factors become decisive:</p><ul><li><p>data quality and concepts/definitions</p></li><li><p>organizational receptiveness</p></li><li><p>learning capacity</p></li><li><p>entrepreneurship and product capability (getting things into production and actually used)</p></li><li><p>clear ownership over time</p></li></ul><p>This is also why &#8220;we ran a pilot&#8221; is rarely the same as &#8220;we&#8217;ve built the capability.&#8221;</p><h3>Think like a venture capitalist and build a learning loop</h3><p>Venture capitalists don&#8217;t invest in a single company and hope it carries everything. They spread risk and give themselves multiple chances.</p><p>A simplified but useful mental model is: out of ten investments, most fail, a few do okay, and one becomes the rocket that pays for the whole thing&#8212;with a strong margin.</p><p>AI efforts are often similar. Outcomes are uncertain, and the biggest wins rarely come from the first attempt. They come from a sequence of initiatives where each step makes the next step better.</p><p>That&#8217;s why AI should be seen as an <strong>investment portfolio with a learning loop</strong>: build, measure, adjust&#8212;without getting stuck because of individual setbacks.</p><p>Uncertainty and learning-loop thinking aren&#8217;t equally easy in all organizations, though. In an organization with a strong <a href="https://www.informerad.se/p/why-ai-gets-stuck-in-some-organizations">Operational Excellence</a> culture, uncertainty can easily trigger more control: more councils, more review cycles, and more &#8220;safe&#8221; wording. The effect is slow progress&#8212;or getting stuck entirely. To pick up speed, you often need to borrow behaviors from other <a href="https://www.informerad.se/p/why-ai-gets-stuck-in-some-organizations">organizational cultures</a> built for faster learning loops, pace, and decision-making authority.</p><h3>Pilot success is not the same as scalable value</h3><p>A pilot often answers the question: &#8220;Can it work?&#8221; Production answers the question: &#8220;Can it work every day, with real users, real data, and real risk?&#8221;</p><p>Some typical gaps:</p><ul><li><p><strong>Data:</strong> the pilot is built on a subset; reality is full of exceptions and unclear definitions.</p></li><li><p><strong>Integration:</strong> it takes time to connect inputs/outputs, permissions, logging, and workflows.</p></li><li><p><strong>Ownership:</strong> when the project team moves on, it becomes unclear who owns the day-to-day results.</p></li><li><p><strong>Actual usage:</strong> if ways of working and behaviors don&#8217;t change, the value is zero&#8212;even if the tech is good.</p></li></ul><p>This is also why a &#8220;successful pilot&#8221; can sometimes be dangerous: it can create expectations without building the prerequisites for scaling.</p><h3>Combination effects: the whole beats any single project</h3><p>Project logic becomes even more misleading because AI initiatives affect one another.</p><p>An initiative that improves data quality, concept models, and access can feel &#8220;too expensive&#8221; to justify through a single use case&#8212;especially if you require every initiative to carry its full foundational cost on its own.</p><p>But if several future initiatives reuse the same foundational work, the business case changes completely. Then the <strong>portfolio</strong> should carry the investment&#8212;not the first pilot.</p><p>This is also where many miss the point: in a portfolio, &#8220;Project A&#8221; can be what makes &#8220;Project B&#8221; possible and &#8220;Project C&#8221; profitable. Trying to price everything in isolation leads you to reject the very building blocks that make scale possible.</p><h3>The portfolio method: from ideas to a roadmap in four steps</h3><p>Here&#8217;s a concrete, repeatable approach:</p><ol><li><p><strong>Set a threshold.</strong> Decide and quantify the minimum benefit that&#8217;s worth pursuing. Without a threshold, everything is important&#8212;and then nothing gets prioritized.</p></li><li><p><strong>Map bottlenecks without talking about AI.</strong> Where is time, quality, or money being wasted in your core flows? Talk to the people who do the work&#8212;not just the people who report on the work.</p></li><li><p><strong>Filter against the threshold.</strong> Which problems are big enough to belong on the roadmap&#8212;i.e., worth spending capacity, risk-management effort, and change muscle on?</p></li><li><p><strong>Prioritize feasibility first.</strong> Once the benefit is already &#8220;good enough,&#8221; the question becomes: what can we get into production fastest, with reasonable risk and clear ownership?</p></li></ol><p>Mini example: &#8220;Customer service has 18 recurring case types &#8594; 5 are large enough to pass the value threshold &#8594; start with 2 where the data is already well structured.&#8221;</p><h3>Execution: key principles</h3><p>Portfolio governance without execution is just a list. Here are principles that help the portfolio survive&#8212;and start delivering.</p><ul><li><p><strong>Treat AI as a product, not a project.</strong> Assign a lifecycle owner. Decide who is responsible for day-to-day outcomes&#8212;not just delivery. That&#8217;s the difference between &#8220;a pilot we did&#8221; and &#8220;a capability we use.&#8221;</p></li><li><p><strong>Actual usage beats model accuracy.</strong> Ask early: do people use this when no one is watching? Have behaviors changed? Would anyone complain if you turned it off for a day? If the answer is no, the value is almost always zero, even if the model is impressive.</p></li><li><p><strong>Build the boring capabilities early.</strong> They don&#8217;t sound exciting, but they&#8217;re the difference between a demo and production:</p><ul><li><p>data definitions and quality</p></li><li><p>access and traceability</p></li><li><p>ownership (business + tech)</p></li><li><p>monitoring and feedback loops</p></li><li><p>risk, security, and privacy</p></li></ul><p>The point: these investments shouldn&#8217;t be &#8220;forced into&#8221; every single initiative. They should be built as shared assets for the portfolio.</p></li><li><p><strong>Run multiple initiatives&#8212;but with stop criteria.</strong> Portfolio doesn&#8217;t mean &#8220;everything at once.&#8221; It means thoughtful prioritization, where you know what&#8217;s required to continue, pause, or stop. The key is that stopping becomes learning, not shame. Otherwise, the organization stops trying.</p></li></ul><h3>Conclusion</h3><p>Project logic makes AI fragile. Portfolio logic makes AI robust.</p><p>Portfolio thinking does two things at once. It reduces the risk that the first pilot becomes the verdict. And it makes it rational to invest in the shared foundational capabilities that enable scaling.</p><p>It also often requires a cultural shift: from &#8220;we must be certain before we do&#8221; to &#8220;we must learn in a controlled way while we do.&#8221;</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.informerad.se/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Informerad.se! Subscribe for free to receive new posts and support my work.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><p><em>Note! The content on this blog reflects my personal opinions and does not represent my employer. As the publisher, I am not responsible for the comments section. Each commenter is responsible for their own posts.</em></p>]]></content:encoded></item><item><title><![CDATA[Sluta driva AI som projekt - börja driva AI som en investeringsportfölj]]></title><description><![CDATA[Om enskilda AI&#8209;initiativ inte n&#229;r upp till f&#246;rv&#228;ntningarna &#228;r det alltf&#246;r l&#228;tt att dra slutsatsen att &#8220;AI funkar inte hos oss&#8221;. Men det &#228;r oftast styrmodellen som &#228;r problemet.]]></description><link>https://www.informerad.se/p/sluta-driva-ai-som-projekt-borja</link><guid isPermaLink="false">https://www.informerad.se/p/sluta-driva-ai-som-projekt-borja</guid><dc:creator><![CDATA[Johan Åhlén]]></dc:creator><pubDate>Wed, 25 Feb 2026 06:28:54 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!zzCV!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F55d1a9bb-17b6-487f-9491-694bf3510358_1966x1097.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p>M&#229;nga organisationer b&#246;rjar med AI s&#229; h&#228;r: v&#228;lj ett lovande anv&#228;ndningsfall, skriv ett business case, presentera f&#246;r ledningen, k&#246;r en pilot, konstatera att f&#246;rv&#228;ntningarna inte uppn&#229;ddes - och ifr&#229;gas&#228;tt AI&#8209;satsningen.</p><p>Men AI beter sig inte som &#8220;vanliga&#8221; projekt. Utfallet f&#246;r enskilda satsningar &#228;r ofta os&#228;kert och v&#228;rdet &#228;r snedf&#246;rdelat. Dessutom sitter det mesta av v&#228;rdet i det som h&#228;nder runt tekniken: f&#246;r&#228;ndrade arbetss&#228;tt, b&#228;ttre beslut och nya aff&#228;rsm&#246;jligheter.</p><p>Det syns i siffrorna:</p><ul><li><p><strong>PwC:s 2026 Global CEO Survey:</strong> <a href="https://www.pwc.com/gx/en/news-room/press-releases/2026/pwc-2026-global-ceo-survey.html">bara </a><strong><a href="https://www.pwc.com/gx/en/news-room/press-releases/2026/pwc-2026-global-ceo-survey.html">12%</a></strong><a href="https://www.pwc.com/gx/en/news-room/press-releases/2026/pwc-2026-global-ceo-survey.html"> av CEO:s</a> s&#228;ger att AI hittills gett b&#229;de kostnads- och int&#228;ktsnytta, och <strong>56%</strong> ser &#228;nnu ingen signifikant finansiell nytta.</p></li><li><p><strong>BCG:s 2025&#8209;studie (&#246;ver 1 250 f&#246;retag):</strong> <a href="https://media-publications.bcg.com/The-Widening-AI-Value-Gap-Sept-2025.pdf">endast </a><strong><a href="https://media-publications.bcg.com/The-Widening-AI-Value-Gap-Sept-2025.pdf">5%</a></strong> skapar AI&#8209;v&#228;rde &#8220;i skala&#8221;, medan <strong>60%</strong> inte ser n&#229;got materiellt v&#228;rde alls trots investeringar.</p></li><li><p><strong>IDC/Lenovo&#8209;studie (refererad av CIO):</strong> <strong><a href="https://www.cio.com/article/3850763/88-of-ai-pilots-fail-to-reach-production-but-thats-not-all-on-it.html">88%</a></strong><a href="https://www.cio.com/article/3850763/88-of-ai-pilots-fail-to-reach-production-but-thats-not-all-on-it.html"> av AI&#8209;piloter</a> n&#229;r inte bred drifts&#228;ttning. F&#246;r varje <strong>33</strong> piloter &#228;r det bara <strong>4</strong> som tar steget till produktion.</p></li></ul><p>Om du d&#229; driver AI som traditionella projekt blir resultatet l&#228;tt en pilotkyrkog&#229;rd. F&#246;rsta piloterna blir ett &#8220;f&#246;rs&#246;k&#8221; &#8211; och om de inte skalar, blir f&#246;rs&#246;ket samtidigt domslutet: &#8220;AI-satsningen var ingen bra id&#233;&#8221;. </p><p>AI&#8209;r&#229;dgivaren och f&#246;rfattaren Tobias Zwingmann f&#229;ngar precis det h&#228;r m&#246;nstret i sin text om <a href="https://blog.tobiaszwingmann.com/p/ai-roadmaps-beat-ai-projects">Why AI Roadmaps beat AI projects</a>. Han beskriver hur f&#246;rsta budgeten inte brukar vara n&#229;gon st&#246;rre sv&#229;righet att f&#229; - men att satsningen kan d&#246; n&#228;r f&#246;rsta prototypen inte levererar som man hoppats. Hans resonemang ligger n&#228;ra mina egna erfarenheter och &#228;r en viktig utg&#229;ngspunkt f&#246;r den h&#228;r artikeln.</p><p>L&#246;sningen &#228;r att byta styrmodell: AI beh&#246;ver drivas som en investeringsportf&#246;lj, med flera satsningar, gemensamma grundf&#246;rm&#229;gor och en roadmap d&#228;r nedlagda initiativ blir l&#228;rdomar ist&#228;llet f&#246;r misslyckanden.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!zzCV!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F55d1a9bb-17b6-487f-9491-694bf3510358_1966x1097.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!zzCV!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F55d1a9bb-17b6-487f-9491-694bf3510358_1966x1097.png 424w, https://substackcdn.com/image/fetch/$s_!zzCV!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F55d1a9bb-17b6-487f-9491-694bf3510358_1966x1097.png 848w, https://substackcdn.com/image/fetch/$s_!zzCV!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F55d1a9bb-17b6-487f-9491-694bf3510358_1966x1097.png 1272w, https://substackcdn.com/image/fetch/$s_!zzCV!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F55d1a9bb-17b6-487f-9491-694bf3510358_1966x1097.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!zzCV!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F55d1a9bb-17b6-487f-9491-694bf3510358_1966x1097.png" width="1456" height="812" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/55d1a9bb-17b6-487f-9491-694bf3510358_1966x1097.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:812,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:4231784,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://www.informerad.se/i/189018344?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F55d1a9bb-17b6-487f-9491-694bf3510358_1966x1097.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!zzCV!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F55d1a9bb-17b6-487f-9491-694bf3510358_1966x1097.png 424w, https://substackcdn.com/image/fetch/$s_!zzCV!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F55d1a9bb-17b6-487f-9491-694bf3510358_1966x1097.png 848w, https://substackcdn.com/image/fetch/$s_!zzCV!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F55d1a9bb-17b6-487f-9491-694bf3510358_1966x1097.png 1272w, https://substackcdn.com/image/fetch/$s_!zzCV!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F55d1a9bb-17b6-487f-9491-694bf3510358_1966x1097.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image"><svg role="img" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg"><g><title></title><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><h3>N&#228;r f&#246;rsta piloten blir domslutet</h3><p>M&#246;nstret &#228;r ofta f&#246;ruts&#228;gbart.</p><p>En id&#233; visar potential. M&#229;nga ser m&#246;jligheterna. Sen kommer verkligheten: data&#229;tkomst, integration, informationss&#228;kerhet, juridik, f&#246;rvaltning, utbildning och processf&#246;r&#228;ndring.</p><p>N&#228;r detta tar emot &#228;r det l&#228;tt att tolka friktion som facit. Det &#228;r d&#229; det &#228;r avg&#246;rande att ha r&#228;tt styrmodell.</p><p>Om ledningen b&#246;rjar fr&#229;ga &#8220;ska vi verkligen forts&#228;tta?&#8221; har satsningen ofta redan tappat energi. Inte f&#246;r att AI saknar potential, utan f&#246;r att man byggt en styrmodell d&#228;r allt f&#246;r mycket h&#228;ngde p&#229; f&#246;rsta f&#246;rs&#246;ket.</p><p>Slutligen kanske organisationen skaffar Copilot-licenser och tycker att nu har de gjort tillr&#228;ckligt inom AI.</p><h3>Varf&#246;r AI skiljer sig fr&#229;n IT</h3><p>I mycket av IT (det jag kallar <a href="https://www.informerad.se/p/basit-och-strategisk-it-tva-vitt">bas-IT</a>) f&#229;r du effekt genom att inf&#246;ra en teknisk l&#246;sning och rulla ut den. Det &#228;r inte alltid helt enkelt - men ofta hyfsat planeringsbart.</p><p>AI d&#228;remot fungerar mer som verksamhetsutveckling. V&#228;rde uppst&#229;r n&#228;r vardagsbeteenden &#228;ndras, n&#228;r beslut blir b&#228;ttre och n&#228;r fl&#246;den g&#246;rs p&#229; ett nytt s&#228;tt.</p><p>Det g&#246;r att helt andra faktorer blir avg&#246;rande:</p><ul><li><p>datakvalitet och begrepp/definitioner</p></li><li><p>mottaglighet i organisationen</p></li><li><p>l&#228;rf&#246;rm&#229;ga</p></li><li><p>entrepren&#246;rskap och produktf&#246;rm&#229;ga (att f&#229; saker i drift och anv&#228;nda dem)</p></li><li><p>tydligt ansvar &#246;ver tid</p></li></ul><p>Det h&#228;r &#228;r ocks&#229; anledningen till att &#8220;vi gjorde en pilot&#8221; s&#228;llan &#228;r samma sak som &#8220;vi har byggt f&#246;rm&#229;gan&#8221;.</p><h3>T&#228;nk som en riskkapitalist och bygg en l&#228;rloop</h3><p>Riskkapitalister investerar inte i ett enda bolag och hoppas att det ska b&#228;ra allt. De sprider risk och ger sig sj&#228;lva flera chanser.</p><p>En f&#246;renklad men nyttig mental modell &#228;r: av tio investeringar misslyckas de allra flesta, n&#229;gra f&#229; g&#229;r okej och en blir raketen som betalar hela kalaset med god marginal.</p><p>AI&#8209;satsningar &#228;r ofta liknande. Utfallet &#228;r os&#228;kert, och de st&#246;rsta vinsterna kommer s&#228;llan fr&#229;n det f&#246;rsta f&#246;rs&#246;ket. De kommer fr&#229;n en serie satsningar d&#228;r varje steg g&#246;r n&#228;sta steg b&#228;ttre.</p><p>Det &#228;r d&#228;rf&#246;r AI b&#246;r ses som en investeringsportf&#246;lj med en l&#228;rloop: bygg, m&#228;t, justera - utan att fastna p&#229; grund av enskilda motg&#229;ngar.</p><p>Os&#228;kerhet och l&#228;rloopst&#228;nk &#228;r dock inte lika l&#228;tt i alla organisationer. I en organisation med tydlig <a href="https://www.informerad.se/p/varfor-fastnar-ai-i-vissa-organisationer">Operational Excellence</a>-kultur, triggar os&#228;kerheten l&#228;tt mer styrning: fler r&#229;d, fler remisser och fler s&#228;kra formuleringar. Effekten blir att det g&#229;r tr&#246;gt eller fastnar. F&#246;r att f&#229; fart beh&#246;ver man d&#229; l&#229;na beteenden fr&#229;n <a href="https://www.informerad.se/p/varfor-fastnar-ai-i-vissa-organisationer">andra organisationskulturer</a> som &#228;r byggda f&#246;r snabbare l&#228;rloopar, tempo och mandat.</p><h3>Pilotframg&#229;ng &#228;r inte samma sak som skalbart v&#228;rde</h3><p>En pilot svarar ofta p&#229; fr&#229;gan: &#8220;g&#229;r det?&#8221;. Produktion svarar p&#229; fr&#229;gan: &#8220;g&#229;r det varje dag, med riktiga anv&#228;ndare, riktiga data och riktig risk?&#8221;.</p><p>N&#229;gra typiska glapp:</p><ul><li><p><strong>Data:</strong> piloten byggs p&#229; ett urval; verkligheten &#228;r full av undantag och otydliga definitioner.</p></li><li><p><strong>Integration:</strong> det tar tid att koppla in&#8209; och utdata, beh&#246;righeter, loggning och arbetsfl&#246;den.</p></li><li><p><strong>&#196;garskap:</strong> n&#228;r projektgruppen g&#229;r vidare blir det oklart vem som ansvarar f&#246;r resultatet i vardagen.</p></li><li><p><strong>Faktisk anv&#228;ndning:</strong> om arbetss&#228;tt och beteenden inte &#228;ndras blir v&#228;rdet noll, &#228;ven om tekniken &#228;r bra.</p></li></ul><p>Det h&#228;r &#228;r ocks&#229; varf&#246;r en &#8220;lyckad pilot&#8221; ibland &#228;r farlig: den kan skapa f&#246;rv&#228;ntningar utan att ha byggt f&#246;ruts&#228;ttningarna f&#246;r att skala.</p><h3>Kombinationseffekter: totalen sl&#229;r varje enskilt projekt</h3><p>Projektlogik blir extra missvisande eftersom AI&#8209;satsningar p&#229;verkar varandra.</p><p>Ett initiativ som f&#246;rb&#228;ttrar datakvalitet, begreppsmodeller och &#229;tkomst kan k&#228;nnas &#8220;f&#246;r dyrt&#8221; att r&#228;kna hem p&#229; ett enda anv&#228;ndningsfall. S&#228;rskilt om man kr&#228;ver att varje initiativ ska b&#228;ra hela sin egen grundkostnad.</p><p>Men om flera kommande satsningar &#229;teranv&#228;nder samma grundarbete f&#246;r&#228;ndras kalkylen totalt. D&#229; &#228;r det portf&#246;ljen som ska b&#228;ra investeringen &#8211; inte f&#246;rsta piloten.</p><p>Det &#228;r ocks&#229; h&#228;r m&#229;nga missar: i en portf&#246;lj kan &#8220;projekt A&#8221; vara det som g&#246;r &#8220;projekt B&#8221; m&#246;jligt, och &#8220;projekt C&#8221; l&#246;nsamt. Att f&#246;rs&#246;ka priss&#228;tta allt isolerat g&#246;r att man v&#228;ljer bort precis de byggstenar som g&#246;r skalan m&#246;jlig.</p><h3>Portf&#246;ljmetoden: fr&#229;n id&#233;er till roadmap i fyra steg</h3><p>S&#229; h&#228;r kan du g&#246;ra det konkret och repeterbart:</p><ol><li><p><strong>S&#228;tt en tr&#246;skel.</strong> Best&#228;m och kvantifiera minsta nytta som &#228;r v&#228;rt att jaga. Utan tr&#246;skel blir allt viktigt - och d&#229; blir inget prioriterat.</p></li><li><p><strong>Kartl&#228;gg flaskhalsar utan att prata AI.</strong> Var f&#246;rspills tid, kvalitet eller pengar i era k&#228;rnfl&#246;den? Prata med dem som g&#246;r jobbet, inte bara dem som rapporterar om jobbet.</p></li><li><p><strong>Filtrera mot tr&#246;skeln.</strong> Vilka problem &#228;r stora nog att vara roadmap&#8209;material - allts&#229; v&#228;rda att l&#228;gga kapacitet, riskarbete och f&#246;r&#228;ndringskraft p&#229;?</p></li><li><p><strong>Prioritera p&#229; genomf&#246;rbarhet f&#246;rst.</strong> N&#228;r nyttan redan &#228;r &#8220;good enough&#8221; blir fr&#229;gan: vad kan vi f&#229; i drift snabbast, med rimlig risk och tydligt &#228;garskap?</p></li></ol><p>Mini&#8209;exempel: &#8220;Kundtj&#228;nst har 18 &#229;terkommande &#228;rendetyper &#8594; 5 &#228;r stora nog att passera v&#228;rdetr&#246;skeln &#8594; starta med 2 d&#228;r data redan &#228;r v&#228;l strukturerad.&#8221;</p><h3>Exekvering: viktiga principer</h3><p>Portf&#246;ljstyrning utan exekvering blir bara en lista. H&#228;r &#228;r principer som g&#246;r att portf&#246;ljen faktiskt &#246;verlever - och b&#246;rjar ge effekt.</p><ul><li><p><strong>Behandla AI som produkt, inte projekt.</strong> Utse en livscykel&#228;gare. Best&#228;m vem som ansvarar f&#246;r resultatet i vardagen, inte bara f&#246;r leveransen. Det &#228;r skillnaden mellan &#8220;en pilot vi gjorde&#8221; och &#8220;en f&#246;rm&#229;ga vi anv&#228;nder&#8221;.</p></li><li><p><strong>Faktisk anv&#228;ndning sl&#229;r modellprecision.</strong> Fr&#229;ga tidigt: anv&#228;nder m&#228;nniskor det h&#228;r n&#228;r ingen tittar? Har beteenden &#228;ndrats? Skulle n&#229;gon klaga om ni st&#228;ngde av det i en dag? Om svaret &#228;r nej, &#228;r v&#228;rdet n&#228;stan alltid noll, &#228;ven om modellen &#228;r imponerande.</p></li><li><p><strong>Bygg de tr&#229;kiga f&#246;rm&#229;gorna tidigt:</strong> De l&#229;ter inte sp&#228;nnande, men de &#228;r skillnaden mellan demo och drift:</p><ul><li><p>datadefinitioner och kvalitet</p></li><li><p>&#229;tkomst och sp&#229;rbarhet</p></li><li><p>&#228;garskap (verksamhet + teknik)</p></li><li><p>monitorering och feedbackloopar</p></li><li><p>risk, s&#228;kerhet och integritet</p></li></ul><p>Po&#228;ngen: de h&#228;r investeringarna ska inte &#8220;tvingas in&#8221; i varje enskilt initiativ. De ska byggas som gemensamma tillg&#229;ngar f&#246;r portf&#246;ljen.</p></li><li><p><strong>K&#246;r flera satsningar - men med stoppkriterier.</strong> Portf&#246;lj betyder inte &#8220;allt p&#229; en g&#229;ng&#8221;. Det betyder en genomt&#228;nkt prioritering, d&#228;r ni vet vad som kr&#228;vs f&#246;r att forts&#228;tta, pausa och stoppa. Det viktiga &#228;r att stopp blir <em>l&#228;rdom</em>, inte <em>skam</em>. Annars slutar organisationen att prova.</p></li></ul><h3>Sammanfattning</h3><p>Projektlogik g&#246;r AI sk&#246;rt. Portf&#246;ljlogik g&#246;r AI robust.</p><p>Portf&#246;ljt&#228;nket g&#246;r tv&#229; saker samtidigt. Det minskar risken att f&#246;rsta piloten blir domslut. Och det g&#246;r det rationellt att investera i de gemensamma grundf&#246;rm&#229;gorna som g&#246;r att satsningen kan skala.</p><p>Det kr&#228;ver ocks&#229; ofta ett kulturellt skifte: fr&#229;n &#8220;vi m&#229;ste vara s&#228;kra innan vi g&#246;r&#8221; till &#8220;vi m&#229;ste l&#228;ra kontrollerat medan vi g&#246;r&#8221;.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.informerad.se/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Tack f&#246;r att du l&#228;ser! Prenumerera f&#246;r att f&#229; nya inl&#228;gg och st&#246;dja mitt arbete.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><p><em>Obs! Inneh&#229;llet p&#229; denna blogg &#228;r mina personliga &#229;sikter och representerar inte min arbetsgivare. Jag som utgivare ansvarar inte f&#246;r kommentarsf&#228;ltet. Den som kommenterar ansvarar sj&#228;lv f&#246;r det som skrivs.</em></p>]]></content:encoded></item><item><title><![CDATA[Utility IT and Strategic IT: two very different things that often get lumped together as “IT” (and “AI”)]]></title><description><![CDATA[We say &#8220;IT&#8221; as if everything were the same. But the difference between utility IT and strategic IT is so large that governing them the same will almost certainly send you in the wrong direction.]]></description><link>https://www.informerad.se/p/utility-it-and-strategic-it-two-very</link><guid isPermaLink="false">https://www.informerad.se/p/utility-it-and-strategic-it-two-very</guid><dc:creator><![CDATA[Johan Åhlén]]></dc:creator><pubDate>Wed, 25 Feb 2026 06:13:32 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!oEcZ!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa84b176f-6d45-4cc0-8e82-33edd2748d2b_2752x1536.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!oEcZ!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa84b176f-6d45-4cc0-8e82-33edd2748d2b_2752x1536.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!oEcZ!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa84b176f-6d45-4cc0-8e82-33edd2748d2b_2752x1536.png 424w, https://substackcdn.com/image/fetch/$s_!oEcZ!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa84b176f-6d45-4cc0-8e82-33edd2748d2b_2752x1536.png 848w, https://substackcdn.com/image/fetch/$s_!oEcZ!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa84b176f-6d45-4cc0-8e82-33edd2748d2b_2752x1536.png 1272w, https://substackcdn.com/image/fetch/$s_!oEcZ!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa84b176f-6d45-4cc0-8e82-33edd2748d2b_2752x1536.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!oEcZ!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa84b176f-6d45-4cc0-8e82-33edd2748d2b_2752x1536.png" width="1456" height="813" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/a84b176f-6d45-4cc0-8e82-33edd2748d2b_2752x1536.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:813,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:5640818,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://www.informerad.se/i/189104233?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa84b176f-6d45-4cc0-8e82-33edd2748d2b_2752x1536.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!oEcZ!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa84b176f-6d45-4cc0-8e82-33edd2748d2b_2752x1536.png 424w, https://substackcdn.com/image/fetch/$s_!oEcZ!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa84b176f-6d45-4cc0-8e82-33edd2748d2b_2752x1536.png 848w, https://substackcdn.com/image/fetch/$s_!oEcZ!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa84b176f-6d45-4cc0-8e82-33edd2748d2b_2752x1536.png 1272w, https://substackcdn.com/image/fetch/$s_!oEcZ!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa84b176f-6d45-4cc0-8e82-33edd2748d2b_2752x1536.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image"><svg role="img" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg"><g><title></title><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>&#8220;IT&#8221; and &#8220;AI&#8221; have become umbrella terms.</p><p>Under the same label you&#8217;ll find everything from email and printing to systems that, in practice, decide how the business actually works. And now &#8220;AI&#8221; is becoming just as broad an umbrella: everything from a helpful chatbot to extensive agentic workflows.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.informerad.se/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Informerad.se! Subscribe for free to receive new posts and support my work.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><p>When things this different get lumped into the same concept, it&#8217;s easy to start managing everything the same way: the same governance, the same budget logic, the same procurement approach, and the same expectation of predictability. And then people are surprised when some initiatives feel &#8220;easy&#8221; while others turn into multi-year learning journeys.</p><p>This isn&#8217;t a new insight. As far back as 2010, Martin Fowler&#8212;author and &#8220;Chief Scientist&#8221; at Thoughtworks&#8212;wrote about a split he calls the <strong><a href="https://martinfowler.com/bliki/UtilityVsStrategicDichotomy.html">Utility vs Strategic Dichotomy</a></strong>.</p><p>Fowler&#8217;s point is that there are two very different types of IT. You can absolutely imagine more categories (and in practice there&#8217;s often a gray zone), but as a starting point his two groups are unexpectedly useful - especially when discussions otherwise get stuck on words like &#8220;agile,&#8221; &#8220;operations,&#8221; or &#8220;innovation.&#8221;</p><h3>Utility IT: standardized services that just need to work</h3><p>Utility IT is the <em>keep-the-lights-on</em> set of services you need every day: email, intranet, identity/login, printing, meeting tools, client platforms, and networks.</p><p>When they work, you barely notice. When they don&#8217;t, you notice immediately (and it can become very critical).</p><p>Utility IT can be compared to the hotel business. You can have different preferences (location, standard, price, framework agreement), but the basic needs are pretty similar. It should simply work as promised.</p><p>In Fowler&#8217;s language this is &#8220;utility&#8221;&#8212;things most people need and that should just work. He tells an anecdote about someone who said:<br><em>&#8220;Software is like sewage pipes. I want it to work reliably and I don&#8217;t want to know about the details.&#8221;</em></p><p>There&#8217;s also <strong>Utility AI</strong>. Just like Utility IT, it consists of standardized everyday services that can be hugely helpful and time-saving&#8212;such as generative AI assistants (like ChatGPT and Copilot). At the same time, you hardly need to care how they work&#8212;as long as they meet security requirements and legal requirements (especially around personal data).</p><p>This is also where many IT organizations are at their strongest. Utility IT (and Utility AI) rewards standardization, process discipline, and strict decision flows. It&#8217;s exactly the kind of work best suited for operational excellence: doing the same thing consistently, safely, and efficiently&#8212;and doing it a little better every day.</p><h3>Strategic IT (and AI): what differentiates you</h3><p>Which brand is on your computer is rarely decisive. Neither is which hotel you choose to stay at.</p><p>But if you&#8217;re going to buy a new house, that&#8217;s a completely different matter. There are many crucial questions: the neighborhood, whether you should buy or perhaps build a new house. Your choices will have a major impact on your life going forward&#8212;monthly costs, commute times, and not least how much you enjoy living there.</p><p><strong>Strategic IT</strong> (and strategic AI) is the kind of thing that matters a lot&#8212;things that <em>differentiate</em> how the business is run.</p><p>It can be an ERP system, production planning, decision support, order-to-delivery, a customer portal, a pricing engine, or a booking flow that forces staffing and planning to work in a new way. Your choices matter enormously&#8212;just like the choices involved in buying or building a house.</p><p>Strategic IT (and AI) is where things most often go wrong.</p><p>Why? Because strategic IT (and AI) is rarely &#8220;just technology.&#8221; It&#8217;s change: ways of working, roles, responsibilities, trade-offs&#8212;and a lot of things that only become clear when systems meet reality. It requires learning, trade-offs, and often quite a bit of renegotiation of &#8220;how we do things around here.&#8221; And it&#8217;s often exactly the non-technical work that gets underestimated&#8212;usually the biggest part of all.</p><p>Strategic IT requires a holistic approach in a completely different way than Utility IT. That&#8217;s why it also requires leadership engagement. It also requires a different kind of responsiveness and prioritization than Utility IT. The very things that are strengths in Utility IT (standardization, process discipline, etc.) can reduce room for maneuver in Strategic IT&#8212;precisely where flexibility, re-evaluation, and speed are needed the most.</p><h3>AI, data, and advanced analytics: often more strategic than you first think</h3><p>While the vast majority of what falls under &#8220;IT&#8221; is Utility IT (according to Martin Fowler perhaps around 95%), AI is a different story.</p><p>Sure, there are often low-hanging &#8220;AI fruits.&#8221; Generative AI assistants can deliver quick productivity gains in everyday work and in many cases function as Utility AI. But the real value of AI, data, and advanced analytics usually only appears when ways of working, decisions, and behaviors change.</p><p>On top of that, the questions are often more complex than in traditional IT. The legal aspects are harder, accountability is less obvious, and the effects are more indirect. Decisions must be made under uncertainty, and risks must be weighed against alternative risks&#8212;that is, what happens if you <em>don&#8217;t</em> act. Without clear leadership engagement, initiatives risk getting stuck as paper products, pilot projects, or isolated experiments that never reach scale.</p><p>That&#8217;s why the split in AI is rarely 95/5 between utility AI and strategic AI. A significantly larger share of AI work has a strategic character.</p><h3>Risks: different kinds of risk (and therefore different governance)</h3><p>Utility IT risks are often about robustness: availability, recovery, security hygiene, and vendor risks. They&#8217;re often possible to define, measure, and follow up in a fairly straightforward way. There are established concepts and metrics: uptime, incident frequency, recovery time, patch levels, compliance, SLA fulfillment.</p><p>The risks are rarely trivial, but they are largely <strong>operational</strong>. They&#8217;re about something stopping working, working more slowly than it should, or being attacked. They&#8217;re risks that can be reduced through redundancy, standardization, clear processes, and disciplined operations.</p><p>That&#8217;s exactly why utility IT is well suited to governance that emphasizes stability, predictability, and continuous improvement. That doesn&#8217;t mean the risks are small&#8212;but they are a different <em>type</em> of risk than in strategic IT.</p><p>The risks of Strategic IT (and AI) look different. Here it&#8217;s less about whether the system is up or down&#8212;and more about whether you actually succeed with the change. The risks are more <strong>business-related and long-term</strong>: the solution doesn&#8217;t fit the ways of working, it&#8217;s implemented but not used as intended, it changes incentives in unexpected ways, or the effect doesn&#8217;t materialize even though the project is delivered according to plan.</p><p>There can also be major delays, scope creep, drawn-out programs that never quite finish, or an &#8220;eternal ramp-up&#8221; where the benefits are always just around the corner. The consequences are often more indirect and harder to measure: lost efficiency, reduced trust, strategic opportunities that slip away.</p><p>That&#8217;s exactly why strategic IT requires a different kind of governance. Here, service levels and operational metrics aren&#8217;t enough. You need active business engagement, the ability to adjust course along the way, and the ability to weigh risk against alternative risk&#8212;what happens if you don&#8217;t change.</p><h3>Bimodal IT</h3><p>So how do Utility IT and Strategic IT relate to the concept of &#8220;Bimodal IT&#8221;?</p><p>Gartner popularized the concept of <strong>bimodal IT</strong>, which is about handling two parallel modes of working: one optimized for predictability and one optimized for exploration. In other contexts this is also called &#8220;two-speed IT.&#8221;</p><p>The bimodal idea received a lot of criticism. Forbes, for example, called it &#8220;<strong><a href="https://www.forbes.com/sites/jasonbloomberg/2015/09/26/bimodal-it-gartners-recipe-for-disaster/">Gartner&#8217;s Recipe for Disaster</a></strong>,&#8221; partly due to the risk of creating the wrong incentives, organizational lock-ins, and an artificial split between &#8220;fast&#8221; and &#8220;slow&#8221; teams. Martin Fowler, in turn, has been clear that bimodal IT is not the same thing as his split between utility and strategic.</p><p>That difference matters. Utility IT versus Strategic IT is not about &#8220;fast vs stable.&#8221; It&#8217;s not a choice between speed and control. It&#8217;s a perspective on the nature of the work, the types of risks involved, and the governance needs&#8212;on what IT actually is to the business, and therefore how it should be led.</p><p>Another misunderstanding that bimodal thinking can reinforce is that speed and quality are opposites&#8212;that you have to choose between being fast or delivering high quality. For example, <a href="https://dora.dev/guides/dora-metrics/">DORA&#8217;s research</a> in software development suggests the opposite. Organizations that deliver quickly also tend to deliver with high quality. Speed and quality often correlate in high-performing teams. So it&#8217;s not that quality is achieved by slowing down&#8212;on the contrary, research indicates that well-functioning teams often manage both at the same time.</p><h3>It starts by asking the right question</h3><p>We need fewer questions about whether something is &#8220;IT&#8221; or &#8220;AI&#8221; - and more questions about what kind of task we&#8217;re actually facing.</p><p>Utility IT/AI keeps the lights on.<br>Strategic IT/AI determines where we&#8217;re going.</p><p>We need both. But they are not the same thing. And they don&#8217;t operate under the same logic.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.informerad.se/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Informerad.se! Subscribe for free to receive new posts and support my work.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><p><em>Note! The content on this blog reflects my personal opinions and does not represent my employer. As the publisher, I am not responsible for the comments section. Each commenter is responsible for their own posts.</em></p>]]></content:encoded></item><item><title><![CDATA[Why AI Gets Stuck in Some Organizations]]></title><description><![CDATA[AI rarely fails for technical reasons. It fails when it collides with a culture that&#8217;s optimized for stability rather than learning.]]></description><link>https://www.informerad.se/p/why-ai-gets-stuck-in-some-organizations</link><guid isPermaLink="false">https://www.informerad.se/p/why-ai-gets-stuck-in-some-organizations</guid><dc:creator><![CDATA[Johan Åhlén]]></dc:creator><pubDate>Wed, 25 Feb 2026 05:50:14 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!4wFA!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F48c0a9c2-9c48-47e4-9081-48f006c21d68_2400x1792.jpeg" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p>I&#8217;m too often meeting organizations that say they&#8217;re &#8220;investing in AI&#8221; &#8212; but where the investment, in practice, consists of PowerPoints, policies, and proofs of concept. When it&#8217;s time to use AI for real, everything grinds to a halt: alignment across multiple layers, overly cautious legal interpretations, and a risk&#8209;minimization mindset of &#8220;better to refrain than risk doing something wrong.&#8221; The result is paper products and &#8220;organizational complication&#8221; instead of organizational development.</p><p>At the same time, there are organizations that move fast. They test, learn, change how they work, and realize benefits early. So why are some fast and others slow?</p><p>I think a big part of the answer is cultural &#8212; and that it&#8217;s connected to organizations&#8217; <strong>value disciplines</strong>.</p><h3>Three value disciplines (and an important trade&#8209;off)</h3><p>Michael Treacy and Fred Wiersema popularized in the 1990s the idea of three <em>value disciplines</em>: <strong>Customer intimacy</strong>, <strong>Product leadership</strong>, and <strong>Operational excellence</strong>.</p><p>The core of the model is that if you want to be truly strong in one discipline, it requires sacrifices in the others. You have to prioritize.</p><p>It can sound like strategy. In practice, it&#8217;s <strong>culture</strong>: which behaviors are rewarded, which are punished, and what the organization reflexively does when uncertainty arises. What gets rewarded becomes the norm, the norm becomes culture, and culture gets baked into the walls - even when the outside world shifts.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!4wFA!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F48c0a9c2-9c48-47e4-9081-48f006c21d68_2400x1792.jpeg" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!4wFA!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F48c0a9c2-9c48-47e4-9081-48f006c21d68_2400x1792.jpeg 424w, https://substackcdn.com/image/fetch/$s_!4wFA!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F48c0a9c2-9c48-47e4-9081-48f006c21d68_2400x1792.jpeg 848w, https://substackcdn.com/image/fetch/$s_!4wFA!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F48c0a9c2-9c48-47e4-9081-48f006c21d68_2400x1792.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!4wFA!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F48c0a9c2-9c48-47e4-9081-48f006c21d68_2400x1792.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!4wFA!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F48c0a9c2-9c48-47e4-9081-48f006c21d68_2400x1792.jpeg" width="1456" height="1087" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/48c0a9c2-9c48-47e4-9081-48f006c21d68_2400x1792.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1087,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:5884787,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/jpeg&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://www.informerad.se/i/189055707?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F48c0a9c2-9c48-47e4-9081-48f006c21d68_2400x1792.jpeg&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!4wFA!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F48c0a9c2-9c48-47e4-9081-48f006c21d68_2400x1792.jpeg 424w, https://substackcdn.com/image/fetch/$s_!4wFA!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F48c0a9c2-9c48-47e4-9081-48f006c21d68_2400x1792.jpeg 848w, https://substackcdn.com/image/fetch/$s_!4wFA!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F48c0a9c2-9c48-47e4-9081-48f006c21d68_2400x1792.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!4wFA!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F48c0a9c2-9c48-47e4-9081-48f006c21d68_2400x1792.jpeg 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image"><svg role="img" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg"><g><title></title><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><h3>The value disciplines</h3><h4>1) Customer intimacy: staying closest to reality</h4><p><strong>The promise</strong>: &#8220;We understand you better than anyone else and adapt to you.&#8221;</p><p>Customer intimacy is built on accepting variation. Two customers/users don&#8217;t always get the same solution &#8212; that&#8217;s the whole point. You invest in relationship&#8209;building, advisory support, tailoring, onboarding, support, and continuous improvement based on real usage.</p><p><strong>Strengths:</strong> responsiveness, relevance, trust, high impact from what you do.<br><strong>You sacrifice:</strong> pure efficiency and standardization. Tailoring costs money and creates complexity.</p><h4>2) Product leadership: being fastest with what&#8217;s new</h4><p><strong>The promise</strong>: &#8220;We lead the way. You get the best, first.&#8221;</p><p>Product leadership isn&#8217;t about having everything perfectly &#8220;tidy&#8221; in every detail. It&#8217;s about the ability to learn quickly. You build the capability to rapidly test hypotheses, measure and evaluate, make decisions under uncertainty, and iterate. Failures are seen as data &#8212; not shame &#8212; as long as they&#8217;re controlled and lead to learning.</p><p><strong>Strengths:</strong> speed, innovation pace, the ability to pivot quickly when reality changes.<br><strong>You sacrifice:</strong> stability, predictability, and sometimes short&#8209;term efficiency. Innovation is &#8220;waste&#8221; &#8212; until it works. And that&#8217;s the point.</p><h4>3) Operational excellence: being the most reliable and efficient</h4><p><strong>The promise</strong>: &#8220;It works. Every time. Cheap, safe, and with consistent quality.&#8221;</p><p>Operational excellence is built on minimizing variation. If you can do the same thing the same way a thousand times, you get lower cost, fewer errors, and more consistent quality. The organization becomes robust: operations are king, deviations are handled systematically, and change happens in a controlled manner.</p><p><strong>Strengths:</strong> quality, stability, compliance, scalable delivery.<br><strong>You sacrifice:</strong> flexibility and pace of change. To be truly great at operational excellence, you need governance mechanisms: change boards, controls, processes, and solid documentation. All of that is rational &#8212; but it has a side effect: change moves slowly.</p><h3>Why AI &#8220;likes&#8221; two value disciplines &#8212; and clashes with the third</h3><p>Adopting AI is, at its core, a change management issue. You don&#8217;t become &#8220;AI&#8209;driven&#8221; by rolling out a tool. You become AI&#8209;driven when ways of working, decisions, and everyday behaviors change.</p><p>It&#8217;s also made harder by the fact that requirements only become clear once you try, results can be difficult to predict, and risks have to be understood in practice. AI is rarely &#8220;install and done.&#8221;</p><p>That&#8217;s why it often moves faster in cultures that are already built for rapid learning loops (Product leadership) or that have high adaptability and a strong focus on user value (Customer intimacy).</p><p>And that&#8217;s why Operational excellence often has the toughest time: what makes operational excellence strong in day&#8209;to&#8209;day operations makes it slow in change and organizational development.</p><p>Operational excellence tends to turn uncertainty into an argument for more alignment: more stakeholders, more councils, more consultation rounds, more &#8220;safe&#8221; wording.</p><p>The paradox is: the newer and more uncertain something is (like AI), the more governance gets triggered &#8212; and the organization ends up <strong>formally active</strong> (policies, guidelines, PowerPoints) but <strong>practically passive</strong> (ways of working and everyday behaviors don&#8217;t change).</p><p>AI can provide enormous leverage for operational excellence (automation, decision support, quality assurance). But the path there often collides with the culture&#8217;s own reflexes.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!pLmT!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F848040c5-1f23-4042-9822-23a1984e08b6_2400x1792.jpeg" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!pLmT!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F848040c5-1f23-4042-9822-23a1984e08b6_2400x1792.jpeg 424w, https://substackcdn.com/image/fetch/$s_!pLmT!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F848040c5-1f23-4042-9822-23a1984e08b6_2400x1792.jpeg 848w, https://substackcdn.com/image/fetch/$s_!pLmT!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F848040c5-1f23-4042-9822-23a1984e08b6_2400x1792.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!pLmT!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F848040c5-1f23-4042-9822-23a1984e08b6_2400x1792.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!pLmT!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F848040c5-1f23-4042-9822-23a1984e08b6_2400x1792.jpeg" width="1456" height="1087" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/848040c5-1f23-4042-9822-23a1984e08b6_2400x1792.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1087,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:5175976,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/jpeg&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://www.informerad.se/i/189055707?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F848040c5-1f23-4042-9822-23a1984e08b6_2400x1792.jpeg&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!pLmT!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F848040c5-1f23-4042-9822-23a1984e08b6_2400x1792.jpeg 424w, https://substackcdn.com/image/fetch/$s_!pLmT!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F848040c5-1f23-4042-9822-23a1984e08b6_2400x1792.jpeg 848w, https://substackcdn.com/image/fetch/$s_!pLmT!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F848040c5-1f23-4042-9822-23a1984e08b6_2400x1792.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!pLmT!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F848040c5-1f23-4042-9822-23a1984e08b6_2400x1792.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image"><svg role="img" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg"><g><title></title><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><h2>Conclusion</h2><p>It&#8217;s important to say this clearly: <strong>Operational excellence is not &#8220;worse.&#8221;</strong> It&#8217;s often a response to a reality where mistakes are expensive &#8212; especially in banks, universities, healthcare, government agencies, and other critical public services. Their mission isn&#8217;t to be first; it&#8217;s to be reliable.</p><p>Treacy and Wiersema make an important point about value disciplines: you can only prioritize one value discipline, but you also need to become <em>good enough</em> at the others. In practice, that means an operational&#8209;excellence organization that wants to realize AI benefits needs to borrow a bit of behavior from the other disciplines: a dose of product leadership (learning loops, speed, mandate) and a dose of customer intimacy (adapting for customers/users and real needs).</p><p>And that doesn&#8217;t happen by itself. It requires <strong>space, courage, and mandate from the very top</strong> &#8212; and that you ask this question:</p><p><strong>What value discipline do we have &#8212; and how do we run our AI initiative so it can make its way through our own culture without getting stuck in it?</strong></p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.informerad.se/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Informerad.se! Subscribe for free to receive new posts and support my work.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><p><em>Note! The content on this blog reflects my personal opinions and does not represent my employer. As the publisher, I am not responsible for the comments section. Each commenter is responsible for their own posts.</em></p>]]></content:encoded></item><item><title><![CDATA[Bas‑IT och strategisk IT: två vitt skilda saker som ofta klumpas ihop som “IT” (och “AI”)]]></title><description><![CDATA[Vi s&#228;ger &#8220;IT&#8221; som om allt vore samma sak. Men skillnaden mellan h&#229;lla&#8209;lampan&#8209;t&#228;nd (Bas&#8209;IT) och strategisk IT &#228;r s&#229; stor att samma styrning n&#228;stan garanterat leder fel.]]></description><link>https://www.informerad.se/p/basit-och-strategisk-it-tva-vitt</link><guid isPermaLink="false">https://www.informerad.se/p/basit-och-strategisk-it-tva-vitt</guid><dc:creator><![CDATA[Johan Åhlén]]></dc:creator><pubDate>Wed, 18 Feb 2026 15:38:11 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!g5PX!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F389009fd-83e9-4614-b6ae-0966690a00f8_2400x1792.jpeg" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!g5PX!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F389009fd-83e9-4614-b6ae-0966690a00f8_2400x1792.jpeg" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!g5PX!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F389009fd-83e9-4614-b6ae-0966690a00f8_2400x1792.jpeg 424w, https://substackcdn.com/image/fetch/$s_!g5PX!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F389009fd-83e9-4614-b6ae-0966690a00f8_2400x1792.jpeg 848w, https://substackcdn.com/image/fetch/$s_!g5PX!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F389009fd-83e9-4614-b6ae-0966690a00f8_2400x1792.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!g5PX!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F389009fd-83e9-4614-b6ae-0966690a00f8_2400x1792.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!g5PX!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F389009fd-83e9-4614-b6ae-0966690a00f8_2400x1792.jpeg" width="1456" height="1087" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/389009fd-83e9-4614-b6ae-0966690a00f8_2400x1792.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1087,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:656675,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/jpeg&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://www.informerad.se/i/187089009?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F389009fd-83e9-4614-b6ae-0966690a00f8_2400x1792.jpeg&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!g5PX!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F389009fd-83e9-4614-b6ae-0966690a00f8_2400x1792.jpeg 424w, https://substackcdn.com/image/fetch/$s_!g5PX!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F389009fd-83e9-4614-b6ae-0966690a00f8_2400x1792.jpeg 848w, https://substackcdn.com/image/fetch/$s_!g5PX!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F389009fd-83e9-4614-b6ae-0966690a00f8_2400x1792.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!g5PX!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F389009fd-83e9-4614-b6ae-0966690a00f8_2400x1792.jpeg 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image"><svg role="img" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg"><g><title></title><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>&#8220;IT&#8221; och &#8220;AI&#8221; har blivit paraplyord.</p><p>Under samma etikett hamnar allt fr&#229;n e&#8209;post och utskrift till system som i praktiken best&#228;mmer hur verksamheten fungerar. Och nu h&#229;ller &#8220;AI&#8221; p&#229; att bli ett lika brett paraply: allt fr&#229;n en hj&#228;lpsam chatbott till omfattande agentiska fl&#246;den.</p><p>N&#228;r s&#229; vitt skilda saker klumpas ihop i samma begrepp &#228;r det l&#228;tt att b&#246;rja hantera allt p&#229; samma s&#228;tt: samma styrning, samma budgetlogik, samma upphandlingsmetodik och samma f&#246;rv&#228;ntan p&#229; f&#246;ruts&#228;gbarhet. Och sen blir man f&#246;rv&#229;nad n&#228;r vissa initiativ k&#228;nns &#8220;enkla&#8221; och andra blir fler&#229;riga l&#228;rresor.</p><p>Det h&#228;r &#228;r inte en ny insikt. Redan 2010 skrev Martin Fowler &#8211; f&#246;rfattare och &#8220;Chief Scientist&#8221; p&#229; Thoughtworks &#8211; om en uppdelning han kallar <strong><a href="https://martinfowler.com/bliki/UtilityVsStrategicDichotomy.html">Utility vs Strategic Dichotomy</a></strong>.</p><p>Fowlers po&#228;ng &#228;r att det finns tv&#229; helt olika typer av IT. Det g&#229;r absolut att t&#228;nka sig fler kategorier (och i praktiken finns det ofta en gr&#229;zon), men som startpunkt &#228;r hans tv&#229; grupper ov&#228;ntat anv&#228;ndbara &#8211; s&#228;rskilt n&#228;r diskussionen annars fastnar i ord som &#8220;agilt&#8221;, &#8220;f&#246;rvaltning&#8221; eller &#8220;innovation&#8221;.</p><p>I den h&#228;r artikeln &#246;vers&#228;tts Fowlers tv&#229; grupper till <strong>Bas&#8209;IT (utility IT)</strong> och <strong>strategisk IT</strong>.</p><div><hr></div><h2>Bas&#8209;IT: standardiserade tj&#228;nster som bara ska fungera</h2><p>Bas&#8209;IT &#228;r de d&#228;r &#8220;h&#229;lla-lampan-t&#228;nd&#8221;-tj&#228;nsterna som beh&#246;vs varje dag: e&#8209;post, intran&#228;t, identitet/inloggning, utskrift, m&#246;tesverktyg, klientplattformar och n&#228;tverk.</p><p>N&#228;r de fungerar m&#228;rks det knappt. N&#228;r det inte fungerar m&#228;rks det direkt (och kan bli v&#228;ldigt kritiskt).</p><p>Bas&#8209;IT kan liknas vid hotellverksamhet. Man kan ha olika preferenser (l&#228;ge, standard, pris, ramavtal), men grundbehoven &#228;r ganska lika. Det skall bara fungera som utlovat.</p><p>I Fowlers spr&#229;k &#228;r detta &#8220;utility&#8221; &#8211; saker som de flesta beh&#246;ver och som bara skall fungera. Han ber&#228;ttar en anekdot om n&#229;gon som sade: &#8220;<em>software is like sewage pipes, I want it to work reliably and I don&#8217;t want to know about the details</em>&#8221;.</p><p>Det finns ocks&#229; Bas&#8209;AI. Precis som Bas-IT &#228;r det standardiserade tj&#228;nster i vardagen som kan vara till oerh&#246;rd nytta och tidsbesparing, exempelvis generativa AI-assistenter (som ChatGPT och Copilot). Samtidigt beh&#246;ver man knappt bry sig om hur de fungerar. Bara de uppfyller s&#228;kerhetskrav och lagkrav (inte minst kring personuppgifter).</p><p>Det &#228;r ocks&#229; h&#228;r m&#229;nga IT&#8209;organisationer &#228;r som starkast. Bas&#8209;IT (och Bas-AI) bel&#246;nar standardisering, processdisciplin och strikta beslutsfl&#246;den. Det &#228;r just den typen av arbete som passar b&#228;st f&#246;r &#8220;operational excellence&#8221;: att g&#246;ra samma sak konsekvent, s&#228;kert och effektivt &#8211; och att g&#246;ra det lite b&#228;ttre varje dag.</p><div><hr></div><h2>Strategisk IT (och AI): det som differentierar</h2><p>Vilket m&#228;rke som st&#229;r p&#229; din dator &#228;r s&#228;llan av avg&#246;rande betydelse. Inte heller vilket hotell du v&#228;ljer att bo p&#229;. </p><p>Om du d&#228;remot skall skaffa nytt hus &#228;r det en helt annan sak. D&#228;r &#228;r m&#229;nga avg&#246;rande fr&#229;gor, s&#229;som bostadsomr&#229;de och om du skall k&#246;pa eller kanske bygga nytt hus. Dina val kommer ha stort betydelse f&#246;r din tillvaro fram&#246;ver, till exempel m&#229;nadskostnader, restider och inte minst trivsel.</p><p>Strategisk IT (och strategisk AI) &#228;r s&#229;dant som har stor betydelse, &#8220;differentierar&#8221;, f&#246;r hur verksamheten drivs.</p><p>Det kan vara aff&#228;rssystem, produktionsplanering, beslutst&#246;dssystem, order&#8209;till&#8209;leverans, kundportal, prismotor, eller ett bokningsfl&#246;de som g&#246;r att bemanning och planering m&#229;ste fungera p&#229; ett nytt s&#228;tt. D&#228;r dina val har mycket stor betydelse, precis som valen n&#228;r du skall k&#246;pa eller bygga hus.</p><p>Det &#228;r oftast i strategisk IT (och AI) det g&#229;r snett.</p><p>Varf&#246;r? F&#246;r att strategisk IT (och AI) s&#228;llan bara &#228;r teknik. Det &#228;r f&#246;r&#228;ndring: arbetss&#228;tt, roller, ansvar, avv&#228;gningar &#8211; och en hel del saker som f&#246;rst blir tydliga n&#228;r systemen m&#246;ter verkligheten. Det kr&#228;ver l&#228;rande, avv&#228;gningar och ofta en hel del omf&#246;rhandling av &#8220;hur vi g&#246;r saker h&#228;r&#8221;. Ofta underskattas just det icke-tekniska arbetet, som brukar vara det allra st&#246;rsta.</p><p>Strategisk IT kr&#228;ver ett helhetst&#228;nk p&#229; helt annat s&#228;tt &#228;n Bas-IT. D&#228;rf&#246;r kr&#228;ver det ocks&#229; ledningens engagemang. Det kr&#228;ver ocks&#229; annan lyh&#246;rdhet och prioriteringar &#228;n Bas-IT. S&#229;dant som &#228;r en styrka inom Bas-IT (standardisering, processdisciplin, mm) kan minska handlingsutrymmet i Strategisk IT just d&#228;r flexibilitet, ompr&#246;vning och tempo beh&#246;vs som allra mest.</p><div><hr></div><h2>AI, data och avancerad analys: ofta mer strategiskt &#228;n man f&#246;rst tror</h2><p>Medan det allra mesta inom &#8220;IT&#8221; &#228;r Bas-IT (enligt Martin Fowler kanske kring 95%), s&#229; &#228;r det en helt annan sak med AI.</p><p>Visst finns det ofta l&#229;gt h&#228;ngande &#8220;AI-frukter&#8221;. Generativa AI-assistenter kan ge snabb produktivitetsvinst i vardagen och fungerar i m&#229;nga fall som Bas-AI. Men den verkliga nyttan av AI, data och avancerad analys uppst&#229;r oftast f&#246;rst n&#228;r arbetss&#228;tt, beslut och beteenden f&#246;r&#228;ndras.</p><p>Dessutom &#228;r fr&#229;gorna ofta mer komplexa &#228;n vid traditionell IT. Juridiken &#228;r sv&#229;rare, ansvarsf&#246;rdelningen mindre sj&#228;lvklar och effekterna mer indirekta. Beslut beh&#246;ver fattas under os&#228;kerhet, och risker m&#229;ste v&#228;gas mot alternativrisker - det vill s&#228;ga vad som h&#228;nder om man avst&#229;r. Utan tydligt ledningsengagemang riskerar initiativen att fastna i pappersprodukter, pilotprojekt eller isolerade experiment som aldrig n&#229;r skala.</p><p>Det &#228;r d&#228;rf&#246;r f&#246;rdelningen inom AI s&#228;llan &#228;r 95/5 mellan Bas-AI och strategisk AI. En betydligt st&#246;rre andel av AI-arbetet har strategisk karakt&#228;r. </p><div><hr></div><h2>Risker: olika riskarter (och d&#228;rf&#246;r olika styrning)</h2><p>Bas-IT-risker handlar ofta om robusthet: tillg&#228;nglighet, &#229;terst&#228;llning, s&#228;kerhetshygien och leverant&#246;rsrisker. De &#228;r ofta m&#246;jliga att definiera, m&#228;ta och f&#246;lja upp p&#229; ett relativt rakt s&#228;tt. Det finns etablerade begrepp och m&#228;tetal: upptid, incidentfrekvens, &#229;terst&#228;llningstid, patchniv&#229;er, regelefterlevnad, SLA-uppfyllnad.</p><p>Riskerna &#228;r s&#228;llan triviala, men de &#228;r i h&#246;g grad <strong>operativa</strong>. De handlar om att n&#229;got slutar fungera, fungerar l&#229;ngsammare &#228;n det borde, eller uts&#228;tts f&#246;r angrepp. Det &#228;r risker som kan minskas genom redundans, standardisering, tydliga processer och disciplin i f&#246;rvaltningen.</p><p>Just d&#228;rf&#246;r l&#228;mpar sig Bas-IT v&#228;l f&#246;r en styrning som betonar stabilitet, f&#246;ruts&#228;gbarhet och kontinuerlig f&#246;rb&#228;ttring. Det betyder inte att riskerna &#228;r sm&#229; &#8211; men de &#228;r av en annan karakt&#228;r &#228;n i strategisk IT.</p><p>Riskerna f&#246;r Strategisk IT (och AI) ser annorlunda ut. H&#228;r handlar det mindre om huruvida systemet &#228;r uppe eller nere - och mer om huruvida man faktiskt lyckas med f&#246;r&#228;ndringen. Riskerna &#228;r i h&#246;gre grad <strong>verksamhetsm&#228;ssiga och l&#229;ngsiktiga</strong>: att l&#246;sningen inte passar arbetss&#228;tten, att den inf&#246;rs men inte anv&#228;nds som t&#228;nkt, att den f&#246;r&#228;ndrar incitament p&#229; ov&#228;ntade s&#228;tt eller att effekten uteblir trots att projektet levereras enligt plan.</p><p>Det kan ocks&#229; bli stora f&#246;rseningar, v&#228;xande scope, utdragna program som aldrig riktigt blir klara, eller en &#8220;evig infasning&#8221; d&#228;r nyttan hela tiden ligger runt n&#228;sta h&#246;rn. Konsekvenserna &#228;r ofta mer indirekta och sv&#229;rare att m&#228;ta: tappad effektivitet, minskat f&#246;rtroende, strategiska m&#246;jligheter som g&#229;r f&#246;rlorade.</p><p>Just d&#228;rf&#246;r kr&#228;ver strategisk IT en annan typ av styrning. H&#228;r r&#228;cker det inte med serviceniv&#229;er och driftm&#229;tt. Det kr&#228;vs aktivt verksamhetsengagemang, en f&#246;rm&#229;ga att justera kursen under resans g&#229;ng och att v&#228;ga risk mot alternativrisk - det vill s&#228;ga vad som h&#228;nder om man inte f&#246;r&#228;ndrar.</p><div><hr></div><h2>Bimodal IT</h2><p>Hur f&#246;rh&#229;ller sig d&#229; Bas-IT och strategisk IT till begreppet &#8220;Bimodal IT&#8221;?</p><p>Gartner gick ut med begreppet <strong>bimodal IT</strong>, som handlar om att hantera tv&#229; sammanh&#228;ngande arbetss&#228;tt: ett som &#228;r optimerat f&#246;r f&#246;ruts&#228;gbarhet och ett som &#228;r optimerat f&#246;r utforskande. Det h&#228;r kallas ocks&#229; &#8220;two&#8209;speed IT&#8221; i andra sammanhang.</p><p>Bimodal&#8209;t&#228;nket fick mycket kritik. Forbes kallade det till exempel &#8220;<strong><a href="https://www.forbes.com/sites/jasonbloomberg/2015/09/26/bimodal-it-gartners-recipe-for-disaster/">Gartner&#8217;s Recipe for Disaster</a></strong>&#8221;, bland annat f&#246;r risken att skapa fel incitament, organisatoriska l&#229;sningar och en konstlad uppdelning mellan &#8220;snabba&#8221; och &#8220;l&#229;ngsamma&#8221; team. Martin Fowler har i sin tur varit tydlig med att bimodal IT inte &#228;r samma sak som hans uppdelning mellan utility och strategic. </p><p>Skillnaden &#228;r viktig. Bas-IT kontra strategisk IT &#228;r inte en fr&#229;ga om &#8220;snabbt vs stabilt&#8221;. Det &#228;r inte ett val mellan tempo och kontroll. Det &#228;r ett perspektiv p&#229; verksamhetsnatur, riskarter och styrbehov - p&#229; vad IT faktiskt <em>&#228;r</em> f&#246;r verksamheten och d&#228;rf&#246;r hur det b&#246;r ledas.</p><p>Ett annat missf&#246;rst&#229;nd som l&#228;tt f&#246;rst&#228;rks av bimodal-t&#228;nket &#228;r att tempo och kvalitet st&#229;r i motsats till varandra. Att man m&#229;ste v&#228;lja mellan att vara snabb <em>eller </em>h&#229;lla h&#246;g kvalitet. Exempelvis <a href="https://dora.dev/guides/dora-metrics/">DORA:s forskning</a> inom mjukvaruutveckling pekar p&#229; motsatsen. De organisationer som levererar snabbt tenderar ocks&#229; att leverera med h&#246;g kvalitet. Tempo och kvalitet korrelerar ofta hos h&#246;gpresterande team. Det &#228;r allts&#229; inte s&#229; att kvalitet uppn&#229;s genom att sakta ner &#8211; <em>tv&#228;rtom </em>visar forskning att v&#228;lfungerande team ofta klarar b&#229;da samtidigt.</p><div><hr></div><h2>Det b&#246;rjar med att st&#228;lla r&#228;tt fr&#229;ga</h2><p>Vi beh&#246;ver f&#228;rre fr&#229;gor om huruvida n&#229;got &#228;r &#8220;IT&#8221; eller &#8220;AI&#8221; - och fler fr&#229;gor om vilken sorts uppgift vi faktiskt st&#229;r inf&#246;r.</p><p>Bas-IT/AI h&#229;ller lampan t&#228;nd.<br>Strategisk IT/AI avg&#246;r vart vi g&#229;r.</p><p>B&#229;da beh&#246;vs. Men de &#228;r inte samma sak. Och de fungerar inte under samma logik.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.informerad.se/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Tack f&#246;r att du l&#228;ser! Prenumerera f&#246;r att f&#229; nya inl&#228;gg och st&#246;dja mitt arbete.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><p><em>Obs! Inneh&#229;llet p&#229; denna blogg &#228;r mina personliga &#229;sikter och representerar inte min arbetsgivare. Jag som utgivare ansvarar inte f&#246;r kommentarsf&#228;ltet. Den som kommenterar ansvarar sj&#228;lv f&#246;r det som skrivs.</em></p>]]></content:encoded></item><item><title><![CDATA[Varför fastnar AI i vissa organisationer?]]></title><description><![CDATA[AI misslyckas s&#228;llan av tekniska sk&#228;l. Det misslyckas n&#228;r det krockar med en kultur som &#228;r optimerad f&#246;r stabilitet snarare &#228;n l&#228;rande.]]></description><link>https://www.informerad.se/p/varfor-fastnar-ai-i-vissa-organisationer</link><guid isPermaLink="false">https://www.informerad.se/p/varfor-fastnar-ai-i-vissa-organisationer</guid><dc:creator><![CDATA[Johan Åhlén]]></dc:creator><pubDate>Thu, 05 Feb 2026 12:46:53 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!UPCG!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F70acaf44-7e2f-4dde-91f3-0b9328d3eaeb_2400x1792.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p>Jag m&#246;ter allt oftare organisationer som s&#228;ger att de &#8220;satsar p&#229; AI&#8221; &#8211; men d&#228;r satsningen i praktiken best&#229;r av PowerPoints, policies och proof of concepts. N&#228;r det ska anv&#228;ndas p&#229; riktigt tar det stopp: f&#246;rankring i m&#229;nga led, &#246;verf&#246;rsiktig juridiktolkning, riskminimering att &#8220;hellre l&#229;ta bli &#228;n riskera att g&#246;ra fel&#8221;. Resultatet blir pappersprodukter och &#8220;verksamhetsinveckling&#8221; ist&#228;llet f&#246;r verksamhetsutveckling.</p><p>Samtidigt finns organisationer som r&#246;r sig snabbt. De testar, l&#228;r, &#228;ndrar arbetss&#228;tt och f&#229;r ut nytta tidigt. S&#229; varf&#246;r &#228;r vissa snabba och andra l&#229;ngsamma?</p><p>Jag tror en stor del av svaret &#228;r kulturellt &#8211; och att det &#228;r kopplat till organisationers <strong>v&#228;rdediscipliner</strong>.</p><h3>Tre v&#228;rdediscipliner (och en viktig m&#229;lkonflikt)</h3><p>Michael Treacy och Fred Wiersema populariserade p&#229; 1990&#8209;talet id&#233;n om tre <em>v&#228;rdediscipliner</em>: <strong>Customer intimacy</strong>, <strong>Product leadership</strong> och <strong>Operational excellence</strong>.</p><p>K&#228;rnan i modellen &#228;r att om man vill vara riktigt stark i en disciplin kr&#228;ver det uppoffringar i de andra. Man m&#229;ste prioritera.</p><p>Det kan l&#229;ta som strategi. I praktiken &#228;r det <strong>kultur</strong>: vilka beteenden som bel&#246;nas, vilka som bestraffas, och vad organisationen reflexm&#228;ssigt g&#246;r n&#228;r os&#228;kerhet uppst&#229;r. Det som premieras blir norm, norm blir kultur, och kultur sitter i v&#228;ggarna &#8211; &#228;ven n&#228;r omv&#228;rlden skiftar.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!UPCG!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F70acaf44-7e2f-4dde-91f3-0b9328d3eaeb_2400x1792.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!UPCG!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F70acaf44-7e2f-4dde-91f3-0b9328d3eaeb_2400x1792.png 424w, https://substackcdn.com/image/fetch/$s_!UPCG!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F70acaf44-7e2f-4dde-91f3-0b9328d3eaeb_2400x1792.png 848w, https://substackcdn.com/image/fetch/$s_!UPCG!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F70acaf44-7e2f-4dde-91f3-0b9328d3eaeb_2400x1792.png 1272w, https://substackcdn.com/image/fetch/$s_!UPCG!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F70acaf44-7e2f-4dde-91f3-0b9328d3eaeb_2400x1792.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!UPCG!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F70acaf44-7e2f-4dde-91f3-0b9328d3eaeb_2400x1792.png" width="1456" height="1087" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/70acaf44-7e2f-4dde-91f3-0b9328d3eaeb_2400x1792.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1087,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:5884787,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://www.informerad.se/i/186882507?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F70acaf44-7e2f-4dde-91f3-0b9328d3eaeb_2400x1792.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!UPCG!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F70acaf44-7e2f-4dde-91f3-0b9328d3eaeb_2400x1792.png 424w, https://substackcdn.com/image/fetch/$s_!UPCG!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F70acaf44-7e2f-4dde-91f3-0b9328d3eaeb_2400x1792.png 848w, https://substackcdn.com/image/fetch/$s_!UPCG!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F70acaf44-7e2f-4dde-91f3-0b9328d3eaeb_2400x1792.png 1272w, https://substackcdn.com/image/fetch/$s_!UPCG!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F70acaf44-7e2f-4dde-91f3-0b9328d3eaeb_2400x1792.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image"><svg role="img" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg"><g><title></title><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><h3>V&#228;rdedisciplinerna</h3><h4>1) Customer intimacy: att vara n&#228;rmast verkligheten</h4><p><strong>L&#246;ftet:</strong> &#8220;Vi f&#246;rst&#229;r dig b&#228;ttre &#228;n n&#229;gon annan och anpassar oss efter dig.&#8221;</p><p>Customer intimacy bygger p&#229; att <strong>acceptera variation</strong>. Tv&#229; kunder/brukare f&#229;r inte alltid samma l&#246;sning &#8211; det &#228;r sj&#228;lva po&#228;ngen. Man satsar p&#229; relationsbyggande, r&#229;dgivning, anpassning, onboarding, support och kontinuerlig f&#246;rb&#228;ttring utifr&#229;n verklig anv&#228;ndning.</p><p><strong>Styrkor:</strong> lyh&#246;rdhet, relevans, f&#246;rtroende, h&#246;gt genomslag f&#246;r det man g&#246;r.<br><strong>Man offrar:</strong> ren effektivitet och standardisering. Anpassning kostar och skapar komplexitet.</p><h4>2) Product leadership: att vara snabbast med det nya</h4><p><strong>L&#246;ftet:</strong> &#8220;Vi leder utvecklingen. Du f&#229;r det b&#228;sta, f&#246;rst.&#8221;</p><p>Product leadership handlar inte om &#8220;ordning och reda&#8221; i varje detalj, utan om f&#246;rm&#229;gan att <strong>l&#228;ra snabbt</strong>. Man bygger f&#246;rm&#229;gan att snabbt testa hypoteser, m&#228;ta och utv&#228;rdera, fatta beslut under os&#228;kerhet och iterera. Misslyckanden ses som data &#8211; inte skam &#8211; s&#229; l&#228;nge de &#228;r kontrollerade och leder till l&#228;rande.</p><p><strong>Styrkor:</strong> tempo, innovationstakt, f&#246;rm&#229;ga att snabbt skifta riktning n&#228;r verkligheten &#228;ndras.<br><strong>Man offrar:</strong> stabilitet, f&#246;ruts&#228;gbarhet och ibland kortsiktig effektivitet. Innovation &#228;r &#8220;sl&#246;seri&#8221; &#8211; tills den fungerar. Och det &#228;r <em>meningen</em>.</p><h4>3) Operational excellence: att vara mest p&#229;litlig och effektiv</h4><p><strong>L&#246;ftet:</strong> &#8220;Det fungerar. Varje g&#229;ng. Billigt, s&#228;kert och med j&#228;mn kvalitet.&#8221;</p><p>Operational excellence bygger p&#229; att minimera variation. Om man kan g&#246;ra samma sak likadant tusen g&#229;nger f&#229;r man l&#228;gre kostnad, f&#228;rre fel och j&#228;mnare kvalitet. Organisationen blir robust: drift &#228;r kung, avvikelser hanteras systematiskt och f&#246;r&#228;ndring sker kontrollerat.</p><p><strong>Styrkor:</strong> kvalitet, stabilitet, regelefterlevnad, skalbar leverans.<br><strong>Man offrar:</strong> flexibilitet och f&#246;r&#228;ndringstakt. F&#246;r att vara riktigt bra p&#229; Operational excellence beh&#246;vs styrmekanismer: f&#246;r&#228;ndringsr&#229;d, kontroller, processer och gedigen dokumentation. Allt detta &#228;r rationellt &#8211; men har en bieffekt: f&#246;r&#228;ndring g&#229;r tr&#246;gt.</p><h3>Varf&#246;r AI &#8220;gillar&#8221; tv&#229; v&#228;rdediscipliner &#8211; och krockar med den tredje</h3><p>AI-inf&#246;rande &#228;r i grunden en f&#246;r&#228;ndringsfr&#229;ga. Du blir inte &#8220;AI-driven&#8221; av att inf&#246;ra ett verktyg. Du blir det n&#228;r arbetss&#228;tt, beslut och vardagsbeteenden &#228;ndras.</p><p>Dessutom f&#246;rsv&#229;ras det av att krav blir tydliga f&#246;rst n&#228;r man provar, resultat kan vara sv&#229;ra att f&#246;ruts&#228;ga och risker m&#229;ste f&#246;rst&#229;s i praktiken. AI &#228;r s&#228;llan &#8220;installera och klart&#8221;.</p><p>D&#228;rf&#246;r g&#229;r det ofta snabbare i kulturer som redan &#228;r byggda f&#246;r snabba l&#228;rloopar (Product leadership) eller med h&#246;g anpassningsbarhet och fokus p&#229; anv&#228;ndarnytta (Customer intimacy).</p><p>Och d&#228;rf&#246;r f&#229;r Operational excellence ofta det tuffare: det som g&#246;r Operational excellence starkt i drift g&#246;r det tr&#246;gt i f&#246;r&#228;ndring och verksamhetsutveckling.</p><p>Operational excellence tenderar att g&#246;ra os&#228;kerhet till ett argument f&#246;r mer f&#246;rankring: fler intressenter, fler r&#229;d, fler remissrundor, fler &#8220;s&#228;kra&#8221; formuleringar.<br>Paradoxen blir: ju mer nytt och os&#228;kert (som AI), desto mer styrning triggas &#8211; och d&#229; hamnar organisationen i ett l&#228;ge d&#228;r den &#228;r <strong>formellt aktiv</strong> (policies, riktlinjer, PowerPoints) men <strong>praktiskt passiv</strong> (arbetss&#228;tt och vardagsbeteenden f&#246;r&#228;ndras inte).</p><p>AI <em>kan</em> ge enorm h&#228;vst&#229;ng f&#246;r Operational excellence (automatisering, beslutst&#246;d, kvalitetss&#228;kring). Men v&#228;gen dit krockar ofta med sj&#228;lva kulturens reflexer.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!Do3M!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F70e44909-d3de-444f-97ae-4ee0c86744a7_2400x1792.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!Do3M!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F70e44909-d3de-444f-97ae-4ee0c86744a7_2400x1792.png 424w, https://substackcdn.com/image/fetch/$s_!Do3M!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F70e44909-d3de-444f-97ae-4ee0c86744a7_2400x1792.png 848w, https://substackcdn.com/image/fetch/$s_!Do3M!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F70e44909-d3de-444f-97ae-4ee0c86744a7_2400x1792.png 1272w, https://substackcdn.com/image/fetch/$s_!Do3M!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F70e44909-d3de-444f-97ae-4ee0c86744a7_2400x1792.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!Do3M!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F70e44909-d3de-444f-97ae-4ee0c86744a7_2400x1792.png" width="1456" height="1087" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/70e44909-d3de-444f-97ae-4ee0c86744a7_2400x1792.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1087,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:5175976,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://www.informerad.se/i/186882507?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F70e44909-d3de-444f-97ae-4ee0c86744a7_2400x1792.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!Do3M!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F70e44909-d3de-444f-97ae-4ee0c86744a7_2400x1792.png 424w, https://substackcdn.com/image/fetch/$s_!Do3M!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F70e44909-d3de-444f-97ae-4ee0c86744a7_2400x1792.png 848w, https://substackcdn.com/image/fetch/$s_!Do3M!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F70e44909-d3de-444f-97ae-4ee0c86744a7_2400x1792.png 1272w, https://substackcdn.com/image/fetch/$s_!Do3M!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F70e44909-d3de-444f-97ae-4ee0c86744a7_2400x1792.png 1456w" sizes="100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image"><svg role="img" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg"><g><title></title><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><h3>Avslutning</h3><p>Det &#228;r viktigt att s&#228;ga det tydligt: <strong>Operational excellence &#228;r inte &#8220;s&#228;mre&#8221;.</strong> Det &#228;r ofta ett svar p&#229; en verklighet d&#228;r fel &#228;r dyra &#8211; s&#228;rskilt i banker, universitet, sjukv&#229;rd, myndigheter och annan samh&#228;llskritisk verksamhet. Deras uppdrag &#228;r inte att vara f&#246;rst, utan att vara p&#229;litliga.</p><p>Treacy och Wiersema har en viktig po&#228;ng om v&#228;rdedisciplinerna: man kan visserligen bara <strong>prioritera en v&#228;rdedisciplin</strong>, men man beh&#246;ver ocks&#229; bli <strong>tillr&#228;ckligt bra</strong> p&#229; de &#246;vriga. I praktiken betyder det att en operational excellence&#8209;organisation som vill f&#229; ut AI&#8209;nytta beh&#246;ver l&#229;na lite beteenden fr&#229;n de andra disciplinerna: en dos av product leadership (l&#228;rloopar, tempo, mandat) och en dos av customer intimacy (anpassning f&#246;r kunder/brukare och verkliga behov).</p><p>Och det sker inte av sig sj&#228;lv. Det kr&#228;vs <strong>utrymme, mod och mandat fr&#229;n h&#246;gsta ledningsh&#229;ll</strong> &#8212; och att man st&#228;ller sig denna fr&#229;ga:</p><p><strong>Vilken v&#228;rdedisciplin har vi &#8211; och hur driver vi v&#229;r AI&#8209;satsning s&#229; att den kan ta sig igenom v&#229;r egen kultur utan att fastna i den?</strong></p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.informerad.se/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Tack f&#246;r att du l&#228;ser! Prenumerera f&#246;r att f&#229; nya inl&#228;gg och st&#246;dja mitt arbete.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><p>Obs! Inneh&#229;llet p&#229; denna blogg &#228;r mina personliga &#229;sikter och representerar inte min arbetsgivare. Jag som utgivare ansvarar inte f&#246;r kommentarsf&#228;ltet. Den som kommenterar ansvarar sj&#228;lv f&#246;r det som skrivs.</p>]]></content:encoded></item><item><title><![CDATA[Innovation utan entreprenörskap stannar vid idéer]]></title><description><![CDATA[Jag m&#246;ter ofta organisationer d&#228;r id&#233;erna finns men resultaten uteblir. H&#228;r f&#246;rklarar jag varf&#246;r entrepren&#246;rskap och mottaglighet &#228;r avg&#246;rande &#8211; och hur ett enkelt hjul med fyra ekrar g&#246;r skillnad.]]></description><link>https://www.informerad.se/p/innovation-utan-entreprenorskap-stannar</link><guid isPermaLink="false">https://www.informerad.se/p/innovation-utan-entreprenorskap-stannar</guid><dc:creator><![CDATA[Johan Åhlén]]></dc:creator><pubDate>Wed, 19 Nov 2025 14:23:56 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!Yyuk!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff6a59026-f5a6-48c9-b26c-5d0363e743d9_1024x1024.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p><strong>Bra id&#233;er r&#228;cker inte.</strong> I n&#228;stan varje organisation finns m&#228;ngder med smarta id&#233;er i huvudet p&#229; kloka medarbetare. &#196;nd&#229; blir s&#229; f&#229; l&#246;sningar genomf&#246;rda i vardagen. <strong>Varf&#246;r &#228;r det s&#229;?</strong> Ofta beror det p&#229; att <strong>innovation utan entrepren&#246;rskap stannar vid inspiration</strong>, och att <strong>&#228;ven de b&#228;sta initiativen fastnar om organisationen inte &#228;r mottaglig</strong>&#8212;om m&#228;nniskor saknar tid, st&#246;d eller inte ser nyttan i sin vardag. N&#228;r n&#229;gon av dessa delar brister, blir effekten noll.</p><p>Lite f&#246;renklat kan det beskrivas som en formel:</p><blockquote><p><strong>Effekt = Innovation &#215; Entrepren&#246;rskap &#215; Mottaglighet.</strong><br>Om n&#229;gon faktor &#228;r noll, blir effekten noll.</p></blockquote><ul><li><p><strong>Innovation</strong> kan ge fantastiska nya m&#246;jligheter, men utan f&#246;rm&#229;gan att driva f&#246;r&#228;ndringarna blir det l&#228;tt &#8220;uppfinnar&#8209;Jocke&#8221;: kul, men utan genomslag.</p></li><li><p><strong>Entrepren&#246;rskap</strong> handlar i grunden om <strong>f&#246;rm&#229;gan att f&#246;rvandla id&#233;er till handling</strong>: att ta en id&#233;, fokusera p&#229; r&#228;tt saker, fatta beslut n&#228;r andra tvekar och driva processen hela v&#228;gen till ett konkret resultat. Det inneb&#228;r att vara lyh&#246;rd f&#246;r hur verkligheten ser ut, att v&#229;ga prioritera, att l&#246;sa hinder i farten och att f&#246;rhindra projekt att fastna i l&#229;nga utredningar.</p></li><li><p><strong>Mottaglighet </strong>&#8211; viljan och kapaciteten hos organisationen att faktiskt ta till sig n&#229;got nytt &#8211; &#228;r lika avg&#246;rande. Mottagligheten handlar b&#229;de om kulturella och praktiska fr&#229;gor. Om m&#229;lgruppen saknar tid att testa, inte har st&#246;d fr&#229;n chefer, inte &#228;r intresserade eller om innovationen inte passar in i deras arbetsfl&#246;den, d&#229; faller &#228;ven den mest lovande innovation platt.</p></li></ul><p>EU beskriver <strong>entrepren&#246;rskapskompetens</strong> som <a href="https://www.eursc.eu/BasicTexts/2018-09-D-69-en-2.pdf">f&#246;rm&#229;gan att agera p&#229; m&#246;jligheter och id&#233;er och omvandla dem till v&#228;rde f&#246;r andra</a>. <strong>Mottaglighet f&#246;r innovation </strong>&#228;r ocks&#229; ganska ganska v&#228;l studerat. Everett Rogers, som lade grunden till teorin om hur innovationer sprids, visade att det inte bara &#228;r sj&#228;lva innovationen i sig utan &#228;ven hur mottaglig milj&#246;n &#228;r som avg&#246;r spridningen.</p><p>Men &#228;ven n&#228;r innovation, entrepren&#246;rskap och mottaglighet finns p&#229; plats r&#228;cker inte det i sig. F&#246;r att en organisation verkligen ska kunna <em>genomf&#246;ra</em> innovation &#8212; inte bara prata om den &#8212; beh&#246;ver flera andra f&#246;ruts&#228;ttningar vara i balans. Jag brukar t&#228;nka p&#229; det som ett <strong>hjul med fyra ekrar</strong>. Om n&#229;gon eker &#228;r f&#246;r svag eller saknas helt, b&#246;rjar hjulet vingla. Du kan fortfarande rulla &#8212; men det g&#229;r l&#229;ngsamt, hackigt och riskerar att krascha.</p><p>H&#228;r &#228;r ekrarna i ett v&#228;lfungerande innovationshjul:</p><h3>1. Pipeline</h3><p>En innovationsverksamhet beh&#246;ver en levande pipeline d&#228;r id&#233;er prioriteras, testas, m&#228;ts och antingen skalas eller stoppas. En bra pipeline ger framf&#246;rh&#229;llning, struktur och riktning.</p><h3>2. Resurser</h3><p>Innovation kr&#228;ver inte stora team, men det kr&#228;ver <em>dedikerad och kontinuerlig</em> tid, <em>motivation</em> och <em>kompetens</em> som inte st&#228;ndigt rycks in i andra br&#228;nder. N&#228;r personal g&#246;r innovationsarbete &#8220;p&#229; ledig tid&#8221; blir det s&#228;llan utrymme.</p><h3>3. Plattform</h3><p>Med plattform menar jag inte bara teknik, utan den samlade infrastruktur som g&#246;r det m&#246;jligt att &#246;ver huvud taget <em>jobba</em> med innovation. Skall man driva innovation inom AI beh&#246;ver man tillg&#229;ng till data av god kvalitet, effektiva rutiner f&#246;r &#229;tkomst, m&#246;jligheter att experimentera och s&#228;tt att s&#228;tta modeller i produktion. F&#246;r andra typer av innovation kan plattformen vara enklare.</p><h3>4. Ledningens st&#246;d</h3><p>Ledningens st&#246;d &#228;r kanske den mest avg&#246;rande ekern. Utan aktiv uppbackning saknar team mandat att ta beslut, prioritera tid och utmana nuvarande arbetss&#228;tt. Med starkt ledningsst&#246;d blir innovation n&#229;got organisationen <em>ska</em> g&#246;ra &#8212; inte n&#229;got som sker i marginalen.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!Yyuk!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff6a59026-f5a6-48c9-b26c-5d0363e743d9_1024x1024.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!Yyuk!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff6a59026-f5a6-48c9-b26c-5d0363e743d9_1024x1024.png 424w, https://substackcdn.com/image/fetch/$s_!Yyuk!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff6a59026-f5a6-48c9-b26c-5d0363e743d9_1024x1024.png 848w, https://substackcdn.com/image/fetch/$s_!Yyuk!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff6a59026-f5a6-48c9-b26c-5d0363e743d9_1024x1024.png 1272w, https://substackcdn.com/image/fetch/$s_!Yyuk!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff6a59026-f5a6-48c9-b26c-5d0363e743d9_1024x1024.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!Yyuk!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff6a59026-f5a6-48c9-b26c-5d0363e743d9_1024x1024.png" width="1024" height="1024" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/f6a59026-f5a6-48c9-b26c-5d0363e743d9_1024x1024.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1024,&quot;width&quot;:1024,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:763775,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://www.informerad.se/i/179335125?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff6a59026-f5a6-48c9-b26c-5d0363e743d9_1024x1024.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!Yyuk!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff6a59026-f5a6-48c9-b26c-5d0363e743d9_1024x1024.png 424w, https://substackcdn.com/image/fetch/$s_!Yyuk!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff6a59026-f5a6-48c9-b26c-5d0363e743d9_1024x1024.png 848w, https://substackcdn.com/image/fetch/$s_!Yyuk!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff6a59026-f5a6-48c9-b26c-5d0363e743d9_1024x1024.png 1272w, https://substackcdn.com/image/fetch/$s_!Yyuk!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff6a59026-f5a6-48c9-b26c-5d0363e743d9_1024x1024.png 1456w" sizes="100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image"><svg role="img" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg"><g><title></title><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>Ytterligare en eker man skulle kunna l&#228;gga till &#228;r mod. Mod &#228;r en underskattad faktor. Innovation kr&#228;ver beslut trots os&#228;kerhet, f&#246;rm&#229;gan att &#8220;kill your darlings&#8221;, och att ge team mandat att k&#246;ra ig&#229;ng &#228;ven n&#228;r allt inte &#228;r perfekt. Utan mod blir organisationen en plats d&#228;r alla v&#228;ntar p&#229; varandra &#8212; och d&#228;r inget h&#228;nder.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.informerad.se/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Tack f&#246;r att du l&#228;ser! Prenumerera f&#246;r att f&#229; nya inl&#228;gg och st&#246;dja mitt arbete.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><p>Obs! Inneh&#229;llet p&#229; denna blogg &#228;r mina personliga &#229;sikter och representerar inte min arbetsgivare. Jag som utgivare ansvarar inte f&#246;r kommentarsf&#228;ltet. Den som kommenterar ansvarar sj&#228;lv f&#246;r det som skrivs.</p>]]></content:encoded></item><item><title><![CDATA[Trappan: från "Excel hell" till verklig nytta]]></title><description><![CDATA[Att lyckas med AI och datastrategi &#228;r ingen quick fix &#8211; det &#228;r en resa som kr&#228;ver ledningens &#228;garskap och l&#229;ngsiktighet. Trappan beskriver resan i fem steg.]]></description><link>https://www.informerad.se/p/trappan-fran-excel-hell-till-verklig</link><guid isPermaLink="false">https://www.informerad.se/p/trappan-fran-excel-hell-till-verklig</guid><dc:creator><![CDATA[Johan Åhlén]]></dc:creator><pubDate>Sun, 14 Sep 2025 08:16:37 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!05Hk!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5ec27a28-8ce3-4e97-aad2-fcd80c1b4940_3325x1920.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!05Hk!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5ec27a28-8ce3-4e97-aad2-fcd80c1b4940_3325x1920.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!05Hk!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5ec27a28-8ce3-4e97-aad2-fcd80c1b4940_3325x1920.png 424w, https://substackcdn.com/image/fetch/$s_!05Hk!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5ec27a28-8ce3-4e97-aad2-fcd80c1b4940_3325x1920.png 848w, https://substackcdn.com/image/fetch/$s_!05Hk!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5ec27a28-8ce3-4e97-aad2-fcd80c1b4940_3325x1920.png 1272w, https://substackcdn.com/image/fetch/$s_!05Hk!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5ec27a28-8ce3-4e97-aad2-fcd80c1b4940_3325x1920.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!05Hk!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5ec27a28-8ce3-4e97-aad2-fcd80c1b4940_3325x1920.png" width="1456" height="841" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/5ec27a28-8ce3-4e97-aad2-fcd80c1b4940_3325x1920.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:841,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:1475873,&quot;alt&quot;:&quot;Illustration av de fem stegen f&#246;r att bli en AI- och datadriven organisation&quot;,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://www.informerad.se/i/173517286?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5ec27a28-8ce3-4e97-aad2-fcd80c1b4940_3325x1920.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="Illustration av de fem stegen f&#246;r att bli en AI- och datadriven organisation" title="Illustration av de fem stegen f&#246;r att bli en AI- och datadriven organisation" srcset="https://substackcdn.com/image/fetch/$s_!05Hk!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5ec27a28-8ce3-4e97-aad2-fcd80c1b4940_3325x1920.png 424w, https://substackcdn.com/image/fetch/$s_!05Hk!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5ec27a28-8ce3-4e97-aad2-fcd80c1b4940_3325x1920.png 848w, https://substackcdn.com/image/fetch/$s_!05Hk!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5ec27a28-8ce3-4e97-aad2-fcd80c1b4940_3325x1920.png 1272w, https://substackcdn.com/image/fetch/$s_!05Hk!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5ec27a28-8ce3-4e97-aad2-fcd80c1b4940_3325x1920.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image"><svg role="img" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg"><g><title></title><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a><figcaption class="image-caption">Fem steg f&#246;r att bli en AI- och datadriven organisation</figcaption></figure></div><p>Alla organisationer pratar idag om att bli mer datadrivna och anv&#228;nda AI. Men vad betyder det egentligen i praktiken? &#196;r det bara att k&#246;pa ett nytt verktyg, anst&#228;lla n&#229;gra data scientists och sedan r&#228;kna hem aff&#228;rsnyttan? Tyv&#228;rr inte. Min erfarenhet efter &#246;ver 25 &#229;r i olika branscher s&#229;som sjukv&#229;rd, f&#246;rs&#228;kring, industri och transport &#228;r att att resan &#228;r p&#229;fallande lik &#8211; oavsett bransch.</p><p>F&#246;r att beskriva den h&#228;r resan anv&#228;nder jag en modell som jag kallar f&#246;r <strong>&#8220;trappan&#8221;</strong>. Den utg&#229;r fr&#229;n en presentation jag en g&#229;ng tog del av fr&#229;n Gartner, men jag har under &#229;ren utvecklat och anpassat den utifr&#229;n mina egna erfarenheter. Trappan best&#229;r av fem steg, fr&#229;n det f&#246;rsta kaosartade f&#246;rs&#246;ket till en fullt transformativ organisation d&#228;r AI och data &#228;r en naturlig del av allt som sker.</p><div><hr></div><h2>Steg 1: Omedveten &#8211; &#8221;Excel hell&#8221;</h2><p>H&#228;r byggs all analys manuellt och personberoendet &#228;r skyh&#246;gt. Best&#228;llningar g&#246;rs ad hoc och leveras ofta via mail eller USB-minnen. Alla g&#246;r p&#229; sitt eget s&#228;tt, och ingen har n&#229;gon helhetsbild &#246;ver organisationens data eller var den kommer ifr&#229;n. Om n&#229;gon slutar kan viktig kunskap f&#246;rsvinna &#246;ver en natt.</p><p>Excel &#228;r ett inte helt ovanligt f&#246;rekommande verktyg f&#246;r data och analys p&#229; den h&#228;r niv&#229;n.</p><div><hr></div><h2>Steg 2: Opportunistisk &#8211; hoppfulla men spretiga initiativ</h2><p>I detta steg har organisationen skaffat sig analysverktyg, kanske gjort n&#229;gra proof of concepts inom AI, och optimismen &#228;r till en b&#246;rjan h&#246;g. Men ganska snabbt dyker problemen upp. Data visar sig vara inl&#229;st i olika system, det &#228;r sv&#229;rt att kombinera information fr&#229;n olika h&#229;ll, och tilltron till datakvaliteten b&#246;rjar eroderas. Det visar sig ofta att organisationen haft lite f&#246;r h&#246;ga tankar om sin datakvalitet. </p><p><em>I ett st&#246;rre bolag jag arbetat med blev detta steg v&#228;ldigt tydligt. Tv&#229; av cheferna i ledningsgruppen hade st&#228;ndiga diskussioner om KPI:erna (nyckeltalen) verkligen st&#228;mde &#8211; och framf&#246;r allt vem av dem som egentligen presterade b&#228;st. N&#228;r sj&#228;lva dataunderlaget &#228;r os&#228;kert blir debatten om siffrorna viktigare &#228;n arbetet med att f&#246;rb&#228;ttra verksamheten. Detta &#228;r ett typexempel p&#229; vad som h&#228;nder n&#228;r organisationen befinner sig p&#229; det opportunistiska steget.</em></p><p>Arbetet med data, analys och AI drivs ofta i stupr&#246;r (med d&#229;ligt samarbete mellan olika delar av organisationen), finansieringen sker projekt f&#246;r projekt och personliga favoritprojekt g&#229;r f&#246;re det som &#228;r viktigast f&#246;r organisationen som helhet. Juridiska fr&#229;gor kan ocks&#229; l&#228;tt bli till stora hinder i arbetet eftersom det &#228;r otydligt vad som g&#228;ller.</p><div><hr></div><h2>Steg 3: Fokuserad &#8211; ledningen tar &#228;garskap</h2><p>P&#229; denna niv&#229; g&#229;r organisationen fr&#229;n spretiga initiativ till ett mer samlat grepp d&#228;r fr&#229;gor om data, analys och AI lyfts upp p&#229; den h&#246;gsta niv&#229;n. Avg&#246;rande f&#246;r att n&#229; detta steg &#228;r att det finns en drivande person i h&#246;gsta ledningen som tar &#228;garskap och driver fr&#229;gorna med uth&#229;llighet.</p><p>N&#228;r &#228;garskapet lyfts till ledningsniv&#229; blir det m&#246;jligt att s&#228;tta en gemensam riktning och tydliga prioriteringar. I st&#228;llet f&#246;r att olika delar av organisationen driver sina egna projekt, kan initiativen styras mot de omr&#229;den d&#228;r de g&#246;r mest nytta f&#246;r helheten. Detta minskar risken f&#246;r att resurser l&#228;ggs p&#229; personliga favoritprojekt och &#246;kar sannolikheten att investeringar faktiskt bidrar till organisationens &#246;vergripande m&#229;l.</p><p>Ledningens &#228;garskap g&#246;r ocks&#229; skillnad n&#228;r det g&#228;ller juridiska och regulatoriska fr&#229;gor. Lagar och regler kring integritet och s&#228;kerhet &#228;r komplexa fr&#229;gor och avv&#228;gningar som ytterst beh&#246;ver avg&#246;ras av h&#246;gsta ledningen. Avsaknad av ledningsbeslut kan leda till s&#229;v&#228;l stort risktagande som handlingsf&#246;rlamning i organisationen.</p><p>P&#229; denna niv&#229; f&#246;rtydligas ansvaret f&#246;r data i organisationen. Vem &#228;ger datan? Vem &#228;r ansvarig f&#246;r kvaliteten? Vem f&#229;r besluta om tillg&#229;ng och anv&#228;ndning? N&#228;r dessa fr&#229;gor f&#229;r svar b&#246;rjar organisationen f&#229; b&#228;ttre ordning p&#229; sin data och &#246;ka datakvaliteten.</p><div><hr></div><h2>Steg 4: Strategisk &#8211; data blir en tillg&#229;ng p&#229; riktigt</h2><p>N&#228;r organisationen n&#229;r den strategiska niv&#229;n har n&#229;got viktigt h&#228;nt: data betraktas inte l&#228;ngre som en biprodukt av verksamheten, utan som n&#246;dv&#228;ndig del. Det inneb&#228;r att organisationen b&#246;rjar t&#228;nka l&#229;ngsiktigt &#8211; inte bara p&#229; den data som beh&#246;vs idag, utan ocks&#229; p&#229; den data som kan visa sig v&#228;rdefull i framtiden.</p><p>P&#229; den h&#228;r niv&#229;n arbetar organisationen med data och AI i st&#246;rre skala. F&#246;r att lyckas kr&#228;vs effektiva processer f&#246;r att hantera m&#229;nga initiativ parallellt, och tekniska plattformar som kan st&#246;dja b&#229;de snabb innovation och stabil drift. Men minst lika viktigt &#228;r den kulturella f&#246;rflyttningen. V&#228;rdet uppst&#229;r i verksamheten genom f&#246;r&#228;ndrade arbetss&#228;tt, b&#228;ttre beslut och nya m&#246;jligheter.</p><p>Det &#228;r ocks&#229; h&#228;r som innovation blir en central del av arbetet med AI. AI handlar ofta om att testa, utforska och experimentera &#8211; ingen vet p&#229; f&#246;rhand vilka modeller eller insikter som faktiskt kommer att fungera. D&#228;rf&#246;r kr&#228;vs en kultur som till&#229;ter utforskande arbete, d&#228;r det &#228;r legitimt att prova sig fram och ibland misslyckas. N&#228;r organisationen v&#229;gar g&#246;ra detta &#246;ppnas helt nya m&#246;jligheter f&#246;r att anv&#228;nda AI och data, inte bara f&#246;r att effektivisera befintliga processer utan ocks&#229; f&#246;r att skapa nya produkter, tj&#228;nster och arbetss&#228;tt.</p><p>Att n&#229; den strategiska niv&#229;n handlar allts&#229; inte bara om att f&#229; tekniken p&#229; plats. Det handlar lika mycket om att bygga f&#246;rtroende, samarbete och nyfikenhet i hela organisationen. F&#246;rst d&#229; kan data bli en verklig strategisk resurs.</p><div><hr></div><h2>Steg 5: Transformativ &#8211; i organisationens &#8220;DNA&#8221;</h2><p>Det h&#246;gsta steget i trappan &#228;r ocks&#229; det mest sv&#229;rf&#229;ngade. H&#228;r &#228;r data och AI inte l&#228;ngre n&#229;got som organisationen &#8220;anv&#228;nder&#8221; i s&#228;rskilda projekt &#8211; det &#228;r en sj&#228;lvklar del av sj&#228;lva verksamheten. Det &#228;r inbyggt i organisationens &#8220;DNA&#8221; och aff&#228;rsid&#233;.</p><p>En tydlig skillnad p&#229; denna niv&#229; &#228;r att data blir &#8220;demokratiserad&#8221;. Det betyder inte att alla har tillg&#229;ng till allt, men att alla medarbetare kan hitta, f&#246;rst&#229; och anv&#228;nda den data de beh&#246;ver f&#246;r att skapa v&#228;rde i sitt arbete. Fr&#229;gan &#228;r inte l&#228;ngre &#8220;f&#229;r vi tag i datan?&#8221; utan &#8220;hur anv&#228;nder vi den p&#229; b&#228;sta s&#228;tt?&#8221;. Det inneb&#228;r ocks&#229; att organisationen lyckats bygga en kultur d&#228;r datadrivet t&#228;nkande inte upplevs som n&#229;got som kommer &#8220;uppifr&#229;n&#8221;, utan som n&#229;got som &#228;r naturligt f&#246;r alla.</p><p>Det transformativa steget &#228;r inget slutm&#229;l i bem&#228;rkelsen att organisationen &#8220;&#228;r klar&#8221;. Tv&#228;rtom &#228;r det en plats d&#228;r st&#228;ndig utveckling &#228;r normen. Organisationen beh&#246;ver forts&#228;tta r&#246;ra sig snabbt och experimentera, men samtidigt ta ansvar f&#246;r etik, integritet och s&#228;kerhet. F&#246;rm&#229;gan att hela tiden anpassa sig, f&#246;rb&#228;ttra sig och skapa nytt v&#228;rde &#8211; det &#228;r just det som g&#246;r data och AI till en verklig konkurrensf&#246;rdel.</p><div><hr></div><h2>Den sv&#229;ra verkligheten</h2><p>Min erfarenhet &#228;r att de flesta fastnar p&#229; niv&#229; tv&#229;. M&#229;nga organisationer k&#228;mpar h&#228;r och den st&#246;rsta utmaningen &#228;r ofta att ledningen inte har tillr&#228;ckligt stort &#228;garskap, och att investeringar forts&#228;tter att ske projekt f&#246;r projekt utan en l&#229;ngsiktig plan.</p><p>Nyttan p&#229; niv&#229; tv&#229; &#228;r begr&#228;nsad eftersom det inte sker tillr&#228;ckligt stora f&#246;r&#228;ndringar i organisationerna utan de jobbar mest som tidigare fast med b&#228;ttre verktyg. Det stora v&#228;rdet av data och AI uppst&#229;r f&#246;rst n&#228;r det leder till f&#246;r&#228;ndrade arbetss&#228;tt och nya m&#246;jligheter.</p><p>Det senaste &#229;ret har dessutom hajpen kring generativ AI och verktyg som ChatGPT gett extra br&#228;nsle &#229;t intresset f&#246;r AI. M&#229;nga organisationer vill nu snabbt hoppa p&#229; t&#229;get och inf&#246;ra den senaste tekniken. Problemet &#228;r att utan ordning p&#229; data, ansvar och processer blir det mest experiment som inte leder till verkligt v&#228;rde.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.informerad.se/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Tack f&#246;r att du l&#228;ser. Prenumerera h&#228;r!</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><p><em>Obs! Inneh&#229;llet p&#229; denna blogg &#228;r mina personliga &#229;sikter och representerar inte min arbetsgivare. Jag som utgivare ansvarar inte f&#246;r kommentarsf&#228;ltet. Den som kommenterar ansvarar sj&#228;lv f&#246;r det som skrivs.</em></p>]]></content:encoded></item><item><title><![CDATA[Unity Health interview part 3 of 3]]></title><description><![CDATA[In this third interview, Derek Beaton (Director of Advanced Analytics at Unity Health Toronto), shares his experiences of applying data science and advanced analytics to solve healthcare problems.]]></description><link>https://www.informerad.se/p/unity-health-interview-part-3-of</link><guid isPermaLink="false">https://www.informerad.se/p/unity-health-interview-part-3-of</guid><dc:creator><![CDATA[Johan Åhlén]]></dc:creator><pubDate>Sun, 19 Jan 2025 08:27:16 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!Dsef!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1471cba7-1488-4536-9479-96c406c3af75_1000x1000.jpeg" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!Dsef!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1471cba7-1488-4536-9479-96c406c3af75_1000x1000.jpeg" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!Dsef!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1471cba7-1488-4536-9479-96c406c3af75_1000x1000.jpeg 424w, https://substackcdn.com/image/fetch/$s_!Dsef!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1471cba7-1488-4536-9479-96c406c3af75_1000x1000.jpeg 848w, https://substackcdn.com/image/fetch/$s_!Dsef!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1471cba7-1488-4536-9479-96c406c3af75_1000x1000.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!Dsef!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1471cba7-1488-4536-9479-96c406c3af75_1000x1000.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!Dsef!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1471cba7-1488-4536-9479-96c406c3af75_1000x1000.jpeg" width="1000" height="1000" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/1471cba7-1488-4536-9479-96c406c3af75_1000x1000.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1000,&quot;width&quot;:1000,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:66196,&quot;alt&quot;:&quot;Derek Beaton portrait&quot;,&quot;title&quot;:null,&quot;type&quot;:&quot;image/jpeg&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="Derek Beaton portrait" title="Derek Beaton portrait" srcset="https://substackcdn.com/image/fetch/$s_!Dsef!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1471cba7-1488-4536-9479-96c406c3af75_1000x1000.jpeg 424w, https://substackcdn.com/image/fetch/$s_!Dsef!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1471cba7-1488-4536-9479-96c406c3af75_1000x1000.jpeg 848w, https://substackcdn.com/image/fetch/$s_!Dsef!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1471cba7-1488-4536-9479-96c406c3af75_1000x1000.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!Dsef!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1471cba7-1488-4536-9479-96c406c3af75_1000x1000.jpeg 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image"><svg role="img" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg"><g><title></title><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a><figcaption class="image-caption">Derek Beaton (source: Unity Health Toronto)</figcaption></figure></div><p><em>This is the third in a series (other parts: <a href="https://www.informerad.se/p/unity-health-interview-part-1-of">part 1</a> and <a href="https://www.informerad.se/p/unity-health-interview-part-2-of">part 2</a>) of interviews with leaders from Unity Health Toronto, including Muhammad Mamdani (Vice-President of Data Science and Advanced Analytics), Michael Page (Director of AI Commercialization), and Derek Beaton (Director of Advanced Analytics), discussing the organization's pioneering efforts in integrating artificial intelligence into healthcare. They share their strategies, challenges, and successes in deploying AI-driven solutions that improve patient outcomes, enhance operational efficiency, and address real-world clinical problems. In collaboration with clinicians, researchers, and technology partners, Unity Health demonstrates how AI is reshaping the future of healthcare.</em></p><p>Read more about <a href="https://unityhealth.to/about-unity-health/ai-at-unity-health/">AI at Unity Health</a> (unityhealth.to).</p><h4>Derek Beaton, <strong>what data sources are most critical to Unity Health&#8217;s AI projects, and how do you manage the challenges of data quality, consistency, and integration?</strong></h4><p>- Instead of targeting specific data sources, our approach is to focus on the problem at hand and then identify the necessary data sources to address the problem. This could range from medical imaging data for certain diagnostic problems to lab and vitals data for clinical prognostication problems to text notes for treatment-related problems.</p><p>- All sources of data are viewed as being critical given the wide variety of problems that can be tackled with AI</p><p>- We embrace the challenges of data quality, consistency, and integration in real world settings! Generally, we manage this by understanding how and why data may have issues or the differences we usually see between historic vs. real time data. We manage these issues by building our solutions around these challenges.</p><blockquote></blockquote><h4><strong>How do you evaluate the performance of your AI/ML models, and what metrics do you consider most important in a healthcare context?</strong></h4><p>- This kind of question is one of my favorites, because I get to play the role of a good statistician: It depends.</p><p>- We have a lot of models across both classification and regression, as well as optimization. Each type of problem requires careful thought about how to evaluate the performance</p><p>- But regardless of approach, there are always fairly standard metrics to rely on (e.g., precision/recall, various error or fit metrics)</p><p>- For examples, some of our clinical deployments must make conservative judgements. So we focus more on not missing cases (e.g., with negative predictive value, false negative rates). In other cases we have benchmarks of clinician performance. So the goals there are to be as least as good in some ways (e.g., sensitivity) while out-performing in other ways (e.g., positive predictive value)</p><div class="pullquote"><p>&#8220;AI literacy is critical to the successful adoption of AI solutions.&#8221;</p></div><h4><strong>What role do you think explainability plays in AI in healthcare?</strong></h4><p>- My perspective on explainability in AI may be a bit contrarian, but I&#8217;m not convinced it plays as big of a role as some people believe. Explainable AI (XAI) has become a stand-in for trust, yet there are many aspects of healthcare&#8212;and life in general&#8212;that we don't fully understand or can&#8217;t explain, but we still trust. For instance, there are common medications that patients take every day, the mechanisms of which are not fully understood, even by many clinicians. Despite this, these medicines are trusted because they&#8217;ve been rigorously tested, clinically validated, and regulated.</p><p>- In the same way, I believe AI tools don&#8217;t always need to be fully explainable for us to trust and use them, as long as they are valid and reliable and subjected to rigorous validation and oversight. While explainability is valuable for understanding and improving these complex systems, I don&#8217;t think it's always a necessity for successful implementation in healthcare. What matters most is ensuring these AI tools are safe, effective, and reliable in practice.</p><p>- Much of what we work with are proxies of other measures. For example, why a model flags that a patient will return to the emergency department is not necessarily why a patient <em>actually</em> returns to the emergency department. So in practice, explainability can be misleading if misunderstood.</p><p>- But if explainability is important, in many cases the there are other approaches for inference and causality that are much better suited than some XAI approaches today.</p><h4><strong>How do you ensure that AI-driven insights are actionable and understandable to healthcare professionals who may not have a technical background?</strong></h4><p>- AI literacy is critical to the successful adoption of AI solutions. As we integrate clinicians into our process, they gain a progressively deeper understanding of data science, machine learning, and AI adoption through &#8216;real world&#8217; experience</p><p>- We have created a learning environment where not only our data scientists learn clinical concepts but our clinicians learn about data science and AI through real world application; clinicians are often provided learning materials on data science and AI relevant to the problem that is being solved and are encouraged to ask questions to grow their knowledge; similarly, our data scientists are encouraged to ask clinical questions to gain a deeper understanding of the problem that the AI solutions aim to address.</p><h4><strong>How do you keep your team updated with the latest advancements in data science and AI, ensuring that Unity Health remains at the forefront of this rapidly evolving field?</strong></h4><p>- I only mean this in the most positive way&#8212;we are a bunch of nerds. Reading and keeping up with the latest developments in AI in medicine is not as hard as it used to be, the content is very accessible, although the volume has significantly increased.</p><p>- Another side of ensuring we remain at the forefront is sort of the opposite of latest advancements. There are a lot of tried-and-true methods that just work. And there are approaches that are just extremely well-suited to solve certain types of problems. While we keep up-to-date, we do also make sure we&#8217;re well versed in techniques that have been around a long time, in some cases many decades and in fact approaching nearly a century (like many well established regression approaches).</p><p>- That said, we are still not experiencing as many providers actually deploying solutions, so we have many global networks formally and informally keeping in touch with technologists, clinicians and corporations on how they are approaching AI in medicine.</p><h4><strong>What exciting AI projects or initiatives can we expect from Unity Health in the near future?</strong></h4><p>- Where we are trying to move really quickly is into high frequency multimodal AI. If you imagine all of the monitors and devices that are bedside, we are interested in streaming all of these different data types in real time into algorithms that can improve care.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.informerad.se/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Informerad.se! Subscribe for free to receive new posts and support my work.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><p><em>Note! The content on this blog reflects my personal opinions and does not represent my employer. As the publisher, I am not responsible for the comments section. Each commenter is responsible for their own posts.</em></p>]]></content:encoded></item><item><title><![CDATA[Unity Health interview part 2 of 3]]></title><description><![CDATA[In this second interview, Michael Page (Director of AI Commercialization at Unity Health Toronto), shares his experiences of successful AI collaborations.]]></description><link>https://www.informerad.se/p/unity-health-interview-part-2-of</link><guid isPermaLink="false">https://www.informerad.se/p/unity-health-interview-part-2-of</guid><dc:creator><![CDATA[Johan Åhlén]]></dc:creator><pubDate>Wed, 15 Jan 2025 07:29:39 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!7eWV!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67842927-2f8e-451b-a9d8-55a8412803e6_3072x1536.jpeg" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!7eWV!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67842927-2f8e-451b-a9d8-55a8412803e6_3072x1536.jpeg" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!7eWV!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67842927-2f8e-451b-a9d8-55a8412803e6_3072x1536.jpeg 424w, https://substackcdn.com/image/fetch/$s_!7eWV!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67842927-2f8e-451b-a9d8-55a8412803e6_3072x1536.jpeg 848w, https://substackcdn.com/image/fetch/$s_!7eWV!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67842927-2f8e-451b-a9d8-55a8412803e6_3072x1536.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!7eWV!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67842927-2f8e-451b-a9d8-55a8412803e6_3072x1536.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!7eWV!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67842927-2f8e-451b-a9d8-55a8412803e6_3072x1536.jpeg" width="1456" height="728" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/67842927-2f8e-451b-a9d8-55a8412803e6_3072x1536.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:728,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:459172,&quot;alt&quot;:&quot;Michael Page portrait&quot;,&quot;title&quot;:null,&quot;type&quot;:&quot;image/jpeg&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="Michael Page portrait" title="Michael Page portrait" srcset="https://substackcdn.com/image/fetch/$s_!7eWV!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67842927-2f8e-451b-a9d8-55a8412803e6_3072x1536.jpeg 424w, https://substackcdn.com/image/fetch/$s_!7eWV!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67842927-2f8e-451b-a9d8-55a8412803e6_3072x1536.jpeg 848w, https://substackcdn.com/image/fetch/$s_!7eWV!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67842927-2f8e-451b-a9d8-55a8412803e6_3072x1536.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!7eWV!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67842927-2f8e-451b-a9d8-55a8412803e6_3072x1536.jpeg 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image"><svg role="img" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg"><g><title></title><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a><figcaption class="image-caption">Michael Page (source: Unity Health Toronto)</figcaption></figure></div><p><em>This is the second in a series (other parts: <a href="https://www.informerad.se/p/unity-health-interview-part-1-of">part 1</a> and <a href="https://www.informerad.se/p/unity-health-interview-part-3-of">part 3</a>) of interviews with leaders from Unity Health Toronto, including Muhammad Mamdani (Vice-President of Data Science and Advanced Analytics), Michael Page (Director of AI Commercialization), and Derek Beaton (Director of Advanced Analytics), discussing the organization's pioneering efforts in integrating artificial intelligence into healthcare. They share their strategies, challenges, and successes in deploying AI-driven solutions that improve patient outcomes, enhance operational efficiency, and address real-world clinical problems. In collaboration with clinicians, researchers, and technology partners, Unity Health demonstrates how AI is reshaping the future of healthcare.</em></p><p>Read more about <a href="https://unityhealth.to/about-unity-health/ai-at-unity-health/">AI at Unity Health</a> (unityhealth.to).</p><h4><strong>Michael Page, what role does collaboration with other healthcare organizations, academic institutions, or technology companies play for Unity Health?</strong></h4><p>- Collaboration is not just valuable&#8212;it&#8217;s essential for the broader adoption and integration of AI in healthcare. At Unity Health, a prime example of this is how our data science team works closely with clinicians to address real-world clinical challenges. This deep internal collaboration has been key to our success in deploying more than 50 AI and analytics tools into practice, delivering tangible benefits in patient care.</p><p>- While internal collaboration is critical, achieving greater health system transformation will require expanding these partnerships beyond the walls of individual institutions. Collaborating with other healthcare organizations, academic institutions, and technology companies allows us to pool expertise, and resources, which accelerates innovation and the spread of AI solutions.</p><p>- For instance, partnerships with academic institutions can drive groundbreaking research, while collaborations with technology companies can facilitate access to cutting-edge tools and platforms that may not yet exist in the healthcare sector. These collaborations help us push the boundaries of what AI can achieve and ensure we stay ahead of technological advancements.</p><p>- Moreover, collaboration across the health system through coalition building is critical for scaling AI solutions and achieving systemic change. While much of this happens informally through academic and professional networks, I envision a future where these efforts become more coordinated. Large-scale collaborations between healthcare providers, researchers, and industry players can help overcome challenges like data sharing, regulatory hurdles, and the clinical validation of AI tools.</p><p>- Looking ahead, partnerships will also play a key role in the commercialization of AI tools. Expanding our collaborations globally, such as with AI Sweden, enables us to not only enhance our own capabilities but also contribute to a more integrated and AI-driven global healthcare ecosystem. Ultimately, collaboration will drive the future of AI in healthcare&#8212;helping us turn innovative ideas into impactful, scalable solutions for patients around the world.</p><p>- While collaborations are essential to drive innovation, we carefully consider strategic collaborations that will add value not only to our respective organizations, but to society at large.</p><div class="pullquote"><p>&#8220;We're fortunate that AI, along with research and innovation more broadly, is a high priority across the organization&#8212;from our Board of Directors to the executive team.&#8221;</p></div><h4><strong>How does Unity Health foster a culture of innovation around AI and data science, and what advice would you give to other healthcare organizations looking to adopt similar technologies?</strong></h4><p>- Fostering a culture of innovation around AI and data science at Unity Health begins with leadership. We're fortunate that AI, along with research and innovation more broadly, is a high priority across the organization&#8212;from our Board of Directors to the executive team. This top-down commitment ensures that everyone is engaged in and aligned with our AI strategy.</p><p>- At St. Michael&#8217;s Hospital, one of our key sites, we have a long-standing tradition of clinician-scientists who balance patient care with teaching and research. This culture of intellectual curiosity and the drive to continuously improve healthcare is deeply embedded within the organization. Our AI initiatives are a natural extension of this legacy, allowing us to use cutting-edge tools to tackle modern challenges.</p><p>- Success also plays a pivotal role in fostering a culture of innovation. For example, our AI-driven tool CHARTwatch, which has been shown to reduce unplanned mortality by over 26% according to recently published research in <em>CMAJ</em>, has been a major confidence booster across the hospital. Success stories like this create a "flywheel effect," where each achievement fuels more interest and momentum. This, in turn, inspires more colleagues to think creatively and collaborate with us on developing the next breakthrough&#8212;whether it's a tool for detecting traumatic brain injuries, predicting massive hemorrhages, or reducing the administrative burden on frontline clinicians so they can spend more time with patients.</p><p>- The only challenge we face is capacity. With so many exciting ideas coming from our colleagues, we currently have a two-year waitlist for new AI project intakes. I wish we had double or triple the resources to tackle them all!</p><h4><strong>Could you tell me about a recent AI project that you&#8217;re particularly satisfied with and explain how it has improved patient care or outcomes at Unity Health?</strong></h4><p>- We&#8217;re satisfied by everything we do! We have such a thorough process before we dive in that it helps ensure we have very high quality products for everything we do. Some of work, like CHARTWatch, is a fundamentally important tool. Recently we showed that CHARTWatch brought (and kept) ICU transfers and mortality (considerably) down since its deployment. And we&#8217;re able to see that because of how long it&#8217;s been in place</p><p>- Like noted, we don&#8217;t have any particular favorites. But we do have approaches across so many domains, each reflecting different types of improvements. There are three in particular we like to highlight.</p><ul><li><p>Our ED RN assignment tool takes schedules, qualifications, and prior assignments of RNs in the ED into account, and will generate assignments (to zones or roles) for the staff. It&#8217;s something that takes a person a long time to do, but our optimization tool can make the assignments in about a minute and the lead nurse can make some edits in about another minute or two. It&#8217;s a significant time savings</p></li><li><p>We have a tool running in the hemodialysis unit that highlights patients at high risk for unplanned admissions (e.g., back to the emergency department). We like to highlight this tool for two reasons. The first is that it took <em>a really long time</em> to go from ideation to production. It&#8217;s a particularly challenging problem from data (extracted from dialysis machines) all the way to workflow integration and intervention. When we deployed, it felt particularly satisfying that we could bring it across the finish line. Even better is that we saw a substantial reduction in unplanned readmissions for high risk patients (25-30% drop).</p></li><li><p>Another set of projects have effectively turned into a suite of tools, or really a larger product and platform all in one. Right now we have two medical imaging models for all head CTs that come into the emergency department: one to identify intracranial hemorrhages and the other to flag cases at high risk for urgent neurosurgical intervention. These models send alerts to a large inter-departmental dashboard that shows all imaging orders in real time. We pulled these tools together and into a more harmonized platform (that we also built) with some new technologies and some older and very robust technologies in medical imaging. We&#8217;ve also built in some comprehensive and very advanced monitoring of the imaging models (via other data sources in different systems throughout the hospital). While largely centralized in the emergency department, this suite is one of our only major interdepartmental tools. It provides alerts and situational awareness across the emergency department, radiology/medical imaging, and neurosurgery. We&#8217;re really excited about the next step, which is to bring this to another one of our hospitals. We&#8217;ll make the jump from inter-departmental tools to organization-wide tools. We can notify neurosurgeons at St. Michael&#8217;s that there&#8217;s a case at St. Joseph&#8217;s emergency department.</p></li></ul><h4><strong>What do you see as the most promising AI-driven healthcare solutions that Unity Health can offer to the broader healthcare community?</strong></h4><p>- We have numerous AI-driven healthcare solutions that we believe will add value to the healthcare system (<a href="https://unityhealth.to/about-unity-health/ai-at-unity-health/">link here</a>), ranging from CHARTwatch &#8211; an early warning system that monitors our patients for clinical deterioration and has been shown to decrease unexpected mortality by 26% - to efficiency solutions such as or ED RN Assignment Solution &#8211; an assignment solution that allocates nurses to zones in the emergency department that has reduced human effort by over 80%.</p><p>- Unfortunately, Unity Health Toronto does not have the ability to deploy these solutions elsewhere given our limited resources and bandwidth, but we are working closely with private sector partners on efforts to scale these solutions globally.</p><p>- Though we can&#8217;t always directly provide our solutions, we can offer our experience and expertise. We build and maintain networks for knowledge sharing.</p><h4><strong>How do you ensure that the AI models developed are both accurate and clinically relevant for Unity Health?</strong></h4><p>- Our rigorous project intake process ensures we work on problems that are highly clinically relevant, as they are proposed by front-line clinicians and requires their full engagement through the development and deployment process. Our solutions are &#8216;developed by clinicians for clinicians&#8217;. While many clinicians may not have a deep understanding of AI, they partner with our AI team where they bring the clinical expertise and we bring the technical expertise. Sine our solutions are developed by our clinicians with our team, adoption is usually quite good.</p><p>- Our technical teams ensure model performance meets a certain standard by focusing on problem-specific performance metrics. We also take considerable effort in understanding the &#8216;status quo&#8217; performance to ensure we have a sound understanding of what level of performance our AI solutions must attain in order to be useful. For example, our CHARTwatch solution is a prognostic solution that predicts patient deterioration (i.e. death or ICU transfer). The standard of care prior to the launch of CHARTwatch was clinician judgement. We collected over 3,000 clinician predictions and were able to quantify physician accuracy (e.g. physician were correct less than 1/3 of the time when they felt a patient would deteriorate). This sets the minimum threshold for AI solution performance in order to provide value.</p><p>- We also go one step further: We evaluate and monitor the impact of our solutions. For example, if a tool is meant to reduce unplanned readmissions to the emergency department, we continue to monitor unplanned emergency department visits after deployment. This tells us that tools are, or are not, still working.</p><h4><strong>What advice would you give to healthcare professionals who are hesitant about adopting AI in their practice?</strong></h4><p>- It's natural for healthcare professionals to have reservations about adopting AI, and we've encountered similar concerns within our own hospitals. These concerns typically come from a place of care and responsibility&#8212;they're rooted in a commitment to patient safety, ethics, and efficacy. And that&#8217;s a good thing. It shows that clinicians are thinking critically about the potential impact of new technologies on their patients.</p><p>- Our approach has been to engage directly with those concerns. It&#8217;s important to first assess current clinical performance and determine whether AI can truly add value. We actively include skeptics in our discussions because their perspectives challenge us to think more rigorously. We set clear, measurable goals, test our AI tools extensively, and only deploy them when there is strong evidence of benefit.</p><p>- For healthcare professionals hesitant about AI, my advice would be to see it as part of the continuous improvement process in care. AI isn't about replacing what works but enhancing it. Medicine has always evolved&#8212;adopting AI is just the next step in that evolution. By maintaining high standards and an open mind, we can integrate AI in ways that make healthcare safer, more efficient, and ultimately, better for patients.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.informerad.se/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Informerad.se! Subscribe for free to receive new posts and support my work.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><p><em>Note! The content on this blog reflects my personal opinions and does not represent my employer. As the publisher, I am not responsible for the comments section. Each commenter is responsible for their own posts.</em></p>]]></content:encoded></item><item><title><![CDATA[Unity Health interview part 1 of 3]]></title><description><![CDATA[In this first interview, Muhammad Mamdani (VP of Data Science and Advanced Analytics at Unity Health Toronto), shares his experiences of integrating artificial intelligence into healthcare.]]></description><link>https://www.informerad.se/p/unity-health-interview-part-1-of</link><guid isPermaLink="false">https://www.informerad.se/p/unity-health-interview-part-1-of</guid><dc:creator><![CDATA[Johan Åhlén]]></dc:creator><pubDate>Mon, 13 Jan 2025 06:50:33 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!As-_!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff9d86a50-b7c4-42ad-980f-c63d94c0c8ac_2000x1333.jpeg" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!As-_!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff9d86a50-b7c4-42ad-980f-c63d94c0c8ac_2000x1333.jpeg" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!As-_!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff9d86a50-b7c4-42ad-980f-c63d94c0c8ac_2000x1333.jpeg 424w, https://substackcdn.com/image/fetch/$s_!As-_!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff9d86a50-b7c4-42ad-980f-c63d94c0c8ac_2000x1333.jpeg 848w, https://substackcdn.com/image/fetch/$s_!As-_!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff9d86a50-b7c4-42ad-980f-c63d94c0c8ac_2000x1333.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!As-_!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff9d86a50-b7c4-42ad-980f-c63d94c0c8ac_2000x1333.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!As-_!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff9d86a50-b7c4-42ad-980f-c63d94c0c8ac_2000x1333.jpeg" width="1456" height="970" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/f9d86a50-b7c4-42ad-980f-c63d94c0c8ac_2000x1333.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:970,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:305876,&quot;alt&quot;:&quot;Muhammad Mamdani portrait&quot;,&quot;title&quot;:null,&quot;type&quot;:&quot;image/jpeg&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="Muhammad Mamdani portrait" title="Muhammad Mamdani portrait" srcset="https://substackcdn.com/image/fetch/$s_!As-_!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff9d86a50-b7c4-42ad-980f-c63d94c0c8ac_2000x1333.jpeg 424w, https://substackcdn.com/image/fetch/$s_!As-_!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff9d86a50-b7c4-42ad-980f-c63d94c0c8ac_2000x1333.jpeg 848w, https://substackcdn.com/image/fetch/$s_!As-_!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff9d86a50-b7c4-42ad-980f-c63d94c0c8ac_2000x1333.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!As-_!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff9d86a50-b7c4-42ad-980f-c63d94c0c8ac_2000x1333.jpeg 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image"><svg role="img" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg"><g><title></title><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a><figcaption class="image-caption">Muhammad Mamdani (source: Unity Health Toronto)</figcaption></figure></div><p><em>This is the first in a series (other parts: <a href="https://www.informerad.se/p/unity-health-interview-part-2-of">part 2</a> and <a href="https://www.informerad.se/p/unity-health-interview-part-3-of">part 3</a>) of interviews with leaders from Unity Health Toronto, including Muhammad Mamdani (Vice-President of Data Science and Advanced Analytics), Michael Page (Director of AI Commercialization), and Derek Beaton (Director of Advanced Analytics), discussing the organization's pioneering efforts in integrating artificial intelligence into healthcare. They share their strategies, challenges, and successes in deploying AI-driven solutions that improve patient outcomes, enhance operational efficiency, and address real-world clinical problems. In collaboration with clinicians, researchers, and technology partners, Unity Health demonstrates how AI is reshaping the future of healthcare.</em></p><p>Read more about <a href="https://unityhealth.to/about-unity-health/ai-at-unity-health/">AI at Unity Health</a> (unityhealth.to).</p><h4><strong>Congratulations Muhammad Mamdani on receiving the 2024 Recognition Award from Canada&#8217;s Drug Agency! Could you share what this award signifies for you and Unity Health? How does it reflect your work with integration of AI in healthcare?</strong></h4><p>- It was a tremendous honour to receive the 2024 Recognition Award from Canada&#8217;s Drug Agency. The award highlights the importance of rigorously evaluating health technologies to ensure the healthcare system receives tangible benefits from existing and emerging technologies given its often limited resources.</p><p>- At Unity Health, we are dedicated to carefully identifying genuine healthcare problems amenable to AI, methodically developing AI solutions by working closely with end-users, deploying AI solutions in a responsible manner, and rigorously evaluating the impacts of our AI solutions.</p><p>- Through this extensive process, we are able to better understand the benefits, and sometimes failures, of AI solutions in healthcare.</p><blockquote></blockquote><h4><strong>Can you tell me about Unity Health and your role in the Canadian healthcare system?</strong></h4><p>- Unity Health Toronto is a publicly funded non-profit and charitable healthcare network that serves the Toronto community. Formed in 2017 through the merger of three healthcare organizations: Providence Healthcare, St. Joseph's Health Centre, and St. Michael's Hospital. The network operates several hospitals, clinics, and long-term care facilities across the city of Toronto, providing a wide range of comprehensive healthcare services to patients.</p><ul><li><p><em><strong>St. Michael's Hospital</strong></em>, a Tier-1 trauma centre and academic hospital, is one of the largest and busiest hospitals in the city, and provides acute care services, including emergency care, critical care, and specialized surgery.</p></li><li><p><em><strong>St. Joseph's Health Centre</strong></em> is a community hospital offering a range of acute care and community health services.</p></li><li><p><em><strong>Providence Healthcare</strong></em> focuses on rehabilitation, long term care and complex continuing care.</p></li></ul><p>- Unity Health Toronto is also committed to education and research, with close ties to the University of Toronto. Notably, Unity Health Toronto is an industry leader, being the only hospital network in Canada to declare AI as a core strategic pillar, as well as establishing an applied health AI program.</p><h4><strong>What is your strategy &#8211; is Unity Health mainly an innovator or early adopter within AI? What impact does your choice of strategy have?</strong></h4><p>- With the endorsement of our Board of Directors and Executive Committee, Unity Health is strategically positioned as an AI Innovator. That means we actively create and implement AI solutions. Innovation carries inherent risks, including potential failures and associated costs. However, these risks are outweighed by the significant benefits we are already seeing in how data science is improving patient outcomes and operational efficiency. &nbsp;As one of the few healthcare organizations at the forefront of AI-driven care, we are excited to help shape the future of healthcare. Being an innovator, however, may require substantial investment of resources and a willingness to accept (and pay for) failure on the path to success.</p><p>- We also recognize other innovators in this space and help facilitate Unity Health Toronto&#8217;s ability to use cutting-edge technology created beyond our walls. Our strategy also acknowledges the value of being early adopters of proven technologies developed by other hospitals and private companies. This is an area where we want to improve and lead. When a solution to a shared problem already exists, we be able to implement it swiftly, reducing unnecessary development efforts and maximizing our focus on impact. But adopting new solutions has its own challenges, such as monitoring or verifying the impact of solutions created elsewhere. This dual approach ensures that we remain both leaders in AI innovation and efficient adopters, balancing pioneering efforts with adopting pragmatic, high-value solutions.</p><h4><strong>Could you give me an overview of how you have chosen to organize AI at Unity Health and the organizational units you have? Also, where is your AI organization located in relation to the whole organization of Unity Health?</strong></h4><p>- This is an important question and one that we frequently revisit as AI continues to evolve, especially in healthcare. Given that AI is still relatively new, we&#8217;ve explored various organizational structures to foster innovation.</p><p>- We initially began as a traditional health research group, fueled by a generous $10 million (CAD) donation from the Li Ka Shing Foundation to our St. Michael&#8217;s Hospital Foundation. This philanthropic support is crucial in Canada and played a pivotal role in jumpstarting our AI journey. With this funding, we invested in building out our technical infrastructure, which took several years but enabled us to mature as an organization. Over time, the hospital began seeing substantial impact from our data-driven projects.</p><p>- Recognizing this, our CEO, Dr. Tim Rutledge, and the hospital executive made the strategic decision to formally embed our data science group into hospital operations under the name <strong>"Data Science and Advanced Analytics" (DSAA)</strong>. While we are part of the Information and Technology group, we operate independently, with extensive collaboration in areas like IT security, privacy, risk, and data governance.</p><p>- Our DSAA team is structured into five distinct groups:</p><ul><li><p><strong>Product Management (PM):</strong> This team oversees product lifecycle management, including problem identification and project management. They work closely with clinical teams to lead deployments and handle change management for AI adoption.</p></li><li><p><strong>Data Integration and Governance (DIG):</strong> The DIG team consists of data engineers responsible for critical tasks such as ETL processes from source systems to our own, data acquisition, building data pipelines, maintaining data warehouses, overseeing the infrastructure to develop and deploy these solutions,, and managing privacy and data governance for the data used in our projects.</p></li><li><p><strong>Advanced Analytics: (AA)</strong> a team of data scientists that apply advanced analytical approaches that span statistical, machine learning, and deep learning approaches as well as operations research and optimization to develop data-driven models to solve hospital-wide problems.</p></li><li><p><strong>Product Development (PD):</strong> The PD team builds solutions that tie together data pipelines from the DIG team and models from the AA team. The PD team has expertise in human factors, design, and software development and focuses on front-end development, ensuring that the AI tools we build are user-friendly and meet the needs of our end-users (e.g. clinicians and administrative decision-makers).</p></li><li><p><strong>AI Commercialization:</strong> The newest addition to our team, this group is responsible for expanding our AI tools to other hospitals and seeking alternative funding sources to further our AI advancements in healthcare. Unity Health Toronto has now formally partnered with two start-ups in health AI and has started to work with several private sector partners to co-develop health AI solutions with the intent to scale them globally.</p></li></ul><blockquote></blockquote><div class="pullquote"><p>&#8220;It's not just about building the technology&#8212;it's about embedding AI into the culture and operations of the hospital in a way that enhances patient care and staff workflows.&#8221;</p></div><h4><strong>What are the biggest challenges you face in implementing AI technologies in a healthcare setting?</strong></h4><p>- There isn&#8217;t just one big challenge when implementing AI in healthcare&#8212;it's a complex, multifaceted process. AI in healthcare is still in its infancy when it comes to real-world application. While there&#8217;s tremendous research happening globally, only a small number of healthcare providers are crossing the "innovation chasm" from research to actual AI deployment in clinical settings. This gap presents a range of challenges, as we are among the few pushing these technologies into real-world use with actual patients.</p><p>- Perhaps the foundational challenge is identifying the &#8216;right&#8217; health problem that may be amenable to AI. While the healthcare system is fraught with challenges, many of them may be solved with better processes and/or resource allocation rather than AI solutions. Some relevant problems may not have the necessary quantity or quality of data to enable AI solution development. Other problems may not have community buy-in or engagement to enable successful adoption. Finding the right problem that is amenable to AI with readily available, timely, and high quality data with full end-user engagement can often be challenging.</p><p>- On a related note, it is critical to have strong leadership buy-in and understanding of AI's potential and limitations. This applies to both clinical and administrative leaders, who must be aligned and well-informed to champion AI initiatives. At Unity Health, we've taken steps to address this by running AI education sessions and implementing staff training programs to help both our clinical and non-clinical teams understand how to interact with the AI tools we develop. It's not just about building the technology&#8212;it's about embedding AI into the culture and operations of the hospital in a way that enhances patient care and staff workflows.</p><h4><strong>What role do you think AI will play at Unity Health over the next 5 to 10 years?</strong></h4><p>- We anticipate that AI will play an increasingly strategic role at Unity Health over the next 5 to 10 years. While we firmly believe that healthcare must remain human at its core, AI has the potential to transform many facets of care. Our focus is on how these tools can empower our clinicians and administrators to deliver data-driven, and possibly accelerated, care without losing the essential human touch.</p><p>- Having already deployed more than 50 AI and analytics tools into clinical practice, we've spent considerable time reflecting on our next steps. Our goal is to expand the role of AI and analytics across every aspect of the organization. Moving forward, we foresee substantial growth in our AI team to conduct more cutting-edge research and to streamline the translation of AI innovations into everyday clinical practice.</p><p>- A key priority for us will be strengthening collaborations, particularly with initiatives like GEMINI, to test the scalability of our tools using data that represents a large number of hospitals across Ontario. This will help us ensure that our solutions are robust and impactful on a larger scale.</p><p>- We are also embracing multimodal AI by developing appropriate infrastructure to enable its application and are starting to explore agentic AI for subspecialized tasks and AI-enabled automation in a safe and responsible manner. At the same time, we focus on applied tools and in many cases still favor more &#8220;traditional&#8221; statistical, ML, and optimization tools that are well suited to solve problems</p><p>- Finally, we are committed to deepening our partnerships, both locally and internationally. Our work with AI Sweden and hospitals across Sweden has been a great start. In the future, we aim to create even more AI partnerships, foster the development of AI companies, and expand our solutions to benefit patients not just in Canada but globally.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!K9bo!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F396aecba-81c6-4ea9-9593-4e007e721d38_2856x2142.jpeg" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!K9bo!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F396aecba-81c6-4ea9-9593-4e007e721d38_2856x2142.jpeg 424w, https://substackcdn.com/image/fetch/$s_!K9bo!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F396aecba-81c6-4ea9-9593-4e007e721d38_2856x2142.jpeg 848w, https://substackcdn.com/image/fetch/$s_!K9bo!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F396aecba-81c6-4ea9-9593-4e007e721d38_2856x2142.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!K9bo!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F396aecba-81c6-4ea9-9593-4e007e721d38_2856x2142.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!K9bo!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F396aecba-81c6-4ea9-9593-4e007e721d38_2856x2142.jpeg" width="1456" height="1092" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/396aecba-81c6-4ea9-9593-4e007e721d38_2856x2142.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1092,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:2385160,&quot;alt&quot;:&quot;Unity Health visit at AI Sweden&quot;,&quot;title&quot;:null,&quot;type&quot;:&quot;image/jpeg&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="Unity Health visit at AI Sweden" title="Unity Health visit at AI Sweden" srcset="https://substackcdn.com/image/fetch/$s_!K9bo!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F396aecba-81c6-4ea9-9593-4e007e721d38_2856x2142.jpeg 424w, https://substackcdn.com/image/fetch/$s_!K9bo!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F396aecba-81c6-4ea9-9593-4e007e721d38_2856x2142.jpeg 848w, https://substackcdn.com/image/fetch/$s_!K9bo!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F396aecba-81c6-4ea9-9593-4e007e721d38_2856x2142.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!K9bo!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F396aecba-81c6-4ea9-9593-4e007e721d38_2856x2142.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image"><svg role="img" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg"><g><title></title><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a><figcaption class="image-caption">Unity Health visit at AI Sweden (source: Unity Health Toronto)</figcaption></figure></div><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.informerad.se/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Informerad.se! Subscribe for free to receive new posts and support my work.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><p><em>Note! The content on this blog reflects my personal opinions and does not represent my employer. As the publisher, I am not responsible for the comments section. Each commenter is responsible for their own posts.</em></p>]]></content:encoded></item><item><title><![CDATA[Markus Lingman, the researcher who wants to transform healthcare data into better patient care]]></title><description><![CDATA[Markus Lingman describes, based on his experiences in clinical practice, leadership, and research, how AI and information-driven healthcare can improve patient care.]]></description><link>https://www.informerad.se/p/markus-lingman-the-researcher-who</link><guid isPermaLink="false">https://www.informerad.se/p/markus-lingman-the-researcher-who</guid><dc:creator><![CDATA[Johan Åhlén]]></dc:creator><pubDate>Fri, 10 Jan 2025 13:43:29 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!Kf_A!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0a44adc6-7614-4f30-a161-d7b86335ca58_912x912.jpeg" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!Kf_A!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0a44adc6-7614-4f30-a161-d7b86335ca58_912x912.jpeg" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!Kf_A!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0a44adc6-7614-4f30-a161-d7b86335ca58_912x912.jpeg 424w, https://substackcdn.com/image/fetch/$s_!Kf_A!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0a44adc6-7614-4f30-a161-d7b86335ca58_912x912.jpeg 848w, https://substackcdn.com/image/fetch/$s_!Kf_A!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0a44adc6-7614-4f30-a161-d7b86335ca58_912x912.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!Kf_A!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0a44adc6-7614-4f30-a161-d7b86335ca58_912x912.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!Kf_A!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0a44adc6-7614-4f30-a161-d7b86335ca58_912x912.jpeg" width="912" height="912" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/0a44adc6-7614-4f30-a161-d7b86335ca58_912x912.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:912,&quot;width&quot;:912,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:300501,&quot;alt&quot;:&quot;Markus Lingman portr&#228;tt&quot;,&quot;title&quot;:null,&quot;type&quot;:&quot;image/jpeg&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="Markus Lingman portr&#228;tt" title="Markus Lingman portr&#228;tt" srcset="https://substackcdn.com/image/fetch/$s_!Kf_A!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0a44adc6-7614-4f30-a161-d7b86335ca58_912x912.jpeg 424w, https://substackcdn.com/image/fetch/$s_!Kf_A!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0a44adc6-7614-4f30-a161-d7b86335ca58_912x912.jpeg 848w, https://substackcdn.com/image/fetch/$s_!Kf_A!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0a44adc6-7614-4f30-a161-d7b86335ca58_912x912.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!Kf_A!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0a44adc6-7614-4f30-a161-d7b86335ca58_912x912.jpeg 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image"><svg role="img" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg"><g><title></title><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a><figcaption class="image-caption">Markus Lingman (photo: Ciprian Gorga)</figcaption></figure></div><p><em>Markus Lingman is a medical doctor and researcher whose career spans clinical practice, leadership, academic research and AI. As a specialist in cardiology and internal medicine, he has seen healthcare challenges up close, but also the opportunities and value of the vast amounts of data produced by the healthcare sector. In this interview, he talks about his journey from his student days to leadership roles in Region Halland, and how his interest in AI and information-driven healthcare emerged. We also learn about concrete examples how data analysis and machine learning can improve care pathways and increase patient security</em>&#8212;<em>without compromising integrity and ethical principles. At the same time, the importance of collaboration and shared standards is highlighted as a decisive factor for Sweden to fully realize the potential of information-driven healthcare.</em></p><h4><strong>Could you tell us about your background? What led you into information-driven healthcare and AI?</strong></h4><p>- A large part of my career has been clinical, as a specialist in cardiology and internal medicine, but about 15 years ago I received my first leadership assignment. I became head of the physicians at a fairly large hospital department, and shortly thereafter I took on a role as departmental manager and soon also as area manager (which in some places is called division manager). For the past few years, I have mainly worked in healthcare management on strategic issues. In parallel with my clinical and leadership career, I&#8217;ve also conducted research. I completed my PhD at Sahlgrenska Academy in 2013, where I am still affiliated, and I also hold an adjunct professorship at Halmstad University.</p><p>- The idea of more information-driven healthcare emerged as a result of my three &#8216;legs&#8217;&#8212;clinical work, management, and research. In the early 2000s, healthcare became digitalized. A side effect was that enormous amounts of data started to be generated and stored. Strangely enough, this goldmine of insights wasn&#8217;t used to any great extent. Most of the metrics used for decision-making fit into an Excel spreadsheet, and we still relied primarily on basic arithmetic. I was frustrated that healthcare&#8212;a large, complex, adaptive system&#8212;didn&#8217;t have better ways to make the right decisions about organizational development. The term &#8216;follow-up&#8217; was common, but follow-up is about looking backward. When developing services, you need to look forward and be able to consider multiple factors simultaneously, so you don&#8217;t create indirect negative side effects of the decisions you make.</p><p>- I saw a great need to look beyond organizational boundaries and focus on how things are going for patients in relation to everything we in healthcare do for them. Healthcare is very good at optimizing silos, even if that happens at the expense of deterioration somewhere else in the system, mainly because there is no fact-based overview of the whole. Information-driven healthcare simply means working more fact-based, holistically, and forward-looking (predictively) by making use of all the data that we in healthcare generate in large quantities every day. Among other things, we developed a resource description model called Patient Encounter Costing (PEC) to describe the resource consumption involved in patient management along complete care pathways. We now use the same approach to calculate climate impact. I just coined the term &#8216;information-driven&#8217; to conceptualize and explain this. The idea came from <em>The Institute for Information Driven Medicine</em> at Harvard.</p><p>- In our work to utilize data in healthcare, we quickly ran into a new problem&#8212;it was just as complex and extensive as the healthcare system it was meant to describe. Traditional statistics and &#8216;business intelligence&#8217; were no longer sufficient. There were too many factors to consider, and the relationships were not linear. At the same time, we were fortunate that machine learning became more broadly accessible, although still largely unknown in healthcare. We began collaborating with researchers in the field at <em>Harvard Medical School</em> and eventually at <em>Halmstad University</em>, and together we refined data into information that led to new insights we could start acting on and monitoring for patient-centric outcomes&#8212;the essence of information-driven healthcare. Over time, neural networks also became applicable, and we are now exploring the possibilities of using language models.</p><p>- Thanks to the capabilities we built, our organization today has entirely different conditions for developing healthcare and optimizing care chains in a fact-based manner than we did 10&#8211;15 years ago.</p><h4><strong>You started studying engineering but switched to medicine. What did you take away from the engineering program?</strong></h4><p>- Before studying medicine, I studied math, physics, and programming, but also national economics, corporate economics, and group economics, as well as languages within the Industrial Economics &#8211; International program. I was fortunate enough to work for short periods at European offices of Swedish corporations. Having a grasp of more advanced mathematics has, of course, been a great help when it comes to AI applications and research. Understanding economics and econometrics has been important because resource management is so central in healthcare&#8212;particularly from an ethical perspective in prioritization, where available resources must be used in the way that creates the greatest benefit. Healthcare still has a lot of work to do in this area. It is important to remember that money is just one of many resources in healthcare, and not necessarily the most constrained one.</p><h4><strong>Could you share an example where your work with information-driven healthcare has improved patient outcomes in Region Halland?</strong></h4><p>- Halland currently has fewer hospital beds than other regions. The decrease in beds is essentially due to limited availability of workforce. This reduction would have been much more problematic if we hadn&#8217;t simultaneously been able to work with the entire care chain so that the need for hospital admission was reduced. For example, we used AI to understand, in depth, what leads a person to end up in the emergency department. Many factors turn out to interact with each other, and some can be influenced. With fact-based oversight and predictive models, we have been able to focus on the right measures. The result is that we do not have more readmissions or higher occupancy levels than before, and at the same time, the patients have better outcomes. We&#8217;ve also been able to build better care pathways for specific patient groups. People with heart failure in Halland need much less inpatient care than was previously the case. Working systematically with data over a long period has contributed to Halland often performing well in national quality comparisons.</p><div class="pullquote"><p>&#8220;Finding new ways to deliver healthcare is absolutely critical in light of our future demographics. Increased precision and prediction will be important factors, and these require data. I&#8217;m also convinced that many people would want to benefit from these capabilities once they themselves become ill or risk becoming ill.&#8221;</p></div><h4><strong>How do you view the balance between information-driven healthcare and protecting patient privacy? How have we managed this balance in Sweden?</strong></h4><p>- We have dealt with sensitive information in healthcare for all time, and one can be impressed by how rarely it is misused. The Health and Medical Services Act states that care should be good, meaning it should be safe, effective, knowledge-based, accessible, personalized, and equitable. None of this can be guaranteed without fact-based development, follow-up, and forward-looking analysis. Therefore, the law grants healthcare providers significant opportunities to do precisely this. The challenge is that protective legislation is tied more to organizational structures than to the patient (who receives care across many parts of the fragmented organization). The requirements to comply with privacy-protecting legislation are high and entail several control mechanisms and restrictions. A current issue is that much of the legislation was created in another era and therefore doesn&#8217;t provide guidance on today&#8217;s new questions. It&#8217;s clear that lawmakers, both nationally and at the EU level, want to see more data utilization to achieve healthcare goals and support research. The development of legislation and its interpretation is intense but takes time.</p><p>- In my view, Sweden has been quite cautious in order to protect privacy. However, I am equally convinced that one can protect privacy and utilize data at the same time. But you need to know what you&#8217;re doing and where the boundaries lie. The benefits for the individual and society are enormous. Finding new ways to deliver healthcare is absolutely critical in light of our future demographics. Increased precision and prediction will be important factors, and these require data. I&#8217;m also convinced that many people would want to benefit from these capabilities once they themselves become ill or risk becoming ill. But to maintain the public&#8217;s trust in healthcare, data must be handled with care and in line with the legislator&#8217;s intentions.</p><h4><strong>In 2020, you were named &#8220;AI Swede of the Year,&#8221; partly for your ability &#8220;to make the complex understandable.&#8221; Do we have a tendency to oversimplify when it comes to health issues?</strong></h4><p>- Absolutely. Many people look for simple answers to incredibly complex questions and quick fixes for difficult problems. Often, interventions are made with the best of intentions but without awareness of indirect negative consequences. This is a classic problem in governance based on indicators. It also leads to solving problems locally, even if that creates new problems elsewhere in the system or at a later point in time. With better access to facts and a culture where people can talk about things as they really are, we improve the conditions for tackling problems in the right way.</p><h4><strong>You have worked with American collaborators. What are the biggest differences in working with information-driven healthcare and AI in Sweden compared to the U.S.?</strong></h4><p>- The Swedish and American healthcare systems are very different. The possibilities for healthcare providers to use data are greater in the American system, but it is afflicted by it being tied to reimbursements. In the U.S., healthcare is an industry like any other, operating in a market. That drives certain types of AI solutions. At the same time, the U.S. has an absolutely outstanding research environment at several top universities and islands of clinical excellence. We can learn from and collaborate with them. But the applications must be consistent with Swedish priorities and how we run our healthcare. I do envy the way some leading academic healthcare organizations in the U.S. have been able to integrate research and clinical care in ways that are not feasible in Sweden.</p><p>- On the other hand, there are also enabling advantages in Sweden&#8212;not least the public&#8217;s trust and the equal access to healthcare. What we have in common is that decisions always improve with a stronger factual basis. Fundamentally, we face the same challenges with the rising healthcare needs of society and a limited supply of skilled workers, even though the demographic profile in the U.S. is more favorable than Sweden&#8217;s.</p><h4><strong>How can Sweden get better at information-driven healthcare and creating value from data? What are the most important measures?</strong></h4><p>- The first short answer is probably increased collaboration&#8212;both across organizational and knowledge boundaries. Right now, many people are sitting in their own corners trying to solve the same problems and interpret the same laws. We need more consensus. This could involve agreeing on standards, legal interpretations, and what is &#8216;good enough.&#8217; At the same time, the government has some homework to do because the state agencies that healthcare providers must relate to are far too siloed from one another. Healthcare providers would benefit from more legal guidance.</p><p>- The second short answer is not to be tempted to treat this as an IT issue. Delivering patient-centric, holistic, forward-looking, and fact-based care is a matter of operations, where clinicians, managers, IT specialists, lawyers, (health) economists, and others must be involved in an integrated way.</p><h4><strong>How do you stay up to date in such fast-changing fields as AI and medicine?</strong></h4><p>- I continuously follow about 20 core scientific journals in medicine, AI, patient safety, healthcare quality, and health economics, and read them regularly. That amounts to many articles to review. At the same time, I&#8217;m fortunate to have like-minded and sharp colleagues both in and outside the country who help with monitoring. It&#8217;s very stimulating. When it comes to ethics and law, I think the WHO, the EU, and the OECD regularly produce excellent summaries and guidelines. Of course, I also read some popular media, but it often lags too far behind. Transformer models were introduced in 2017 but only gained widespread media attention in 2024. By then, all the important developments had already happened.</p><h4><strong>What do you see as the most interesting AI trends in healthcare over the next 3&#8211;5 years?</strong></h4><p>- What I&#8217;m most curious about is how we can make use of the enormous amount of unstructured data in healthcare, i.e., free text. It contains much more information than the structured data, and the possibilities to extract value from it have increased dramatically thanks to transformer models (e.g., language models). However, there are a number of uncertainties that we need to clarify before we see major effects in which healthcare professionals can be relieved on a large scale.</p><p>- What I&#8217;m hoping for is that we will be able to apply clinical decision support in day-to-day care to increase diagnostic precision, select tailored treatment more quickly for each individual, and take action based on risk rather than what has already happened. However, this will require clarity on issues of responsibility and what demands we should place on decision support tools. Is it enough if they are as good as an average doctor? What standards should we apply regarding transparency and mitigating bias? Upcoming legislation addresses this, but it needs to be translated into practice and discussed in society. We need to develop a more common view of the balance between risk and benefit when it comes to AI, similar to what we already have for pharmaceuticals.</p><h4><strong>If you could give one piece of advice to young physicians and researchers who want to work with AI, what would it be?</strong></h4><p>- I think the question is a bit misdirected. Sooner or later, a physician or researcher who does not have a decent grasp of AI will become less relevant or even obsolete. We&#8217;re currently working on how AI knowledge should be included in all Swedish medical education. It&#8217;s just a bit unfortunate that it took until 2024. You don&#8217;t need to understand every detail, but you must have a good sense of the risks and limitations, just like any other tool we use to help patients.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.informerad.se/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Informerad.se! Subscribe for free to receive new posts and support my work.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><p><em>Note! The content on this blog reflects my personal opinions and does not represent my employer. As the publisher, I am not responsible for the comments section. Each commenter is responsible for their own posts.</em></p>]]></content:encoded></item><item><title><![CDATA[Markus Lingman, forskaren som vill förvandla vårdens data till bättre patientvård]]></title><description><![CDATA[Markus Lingman beskriver, grundat i hans erfarenheter fr&#229;n klinik, ledarskap och forskning, hur AI och informationsdriven v&#229;rd f&#246;rb&#228;ttrar patienternas v&#229;rdkedjor.]]></description><link>https://www.informerad.se/p/markus-lingman-forskaren-som-vill</link><guid isPermaLink="false">https://www.informerad.se/p/markus-lingman-forskaren-som-vill</guid><dc:creator><![CDATA[Johan Åhlén]]></dc:creator><pubDate>Thu, 09 Jan 2025 06:51:52 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!Kf_A!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0a44adc6-7614-4f30-a161-d7b86335ca58_912x912.jpeg" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!Kf_A!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0a44adc6-7614-4f30-a161-d7b86335ca58_912x912.jpeg" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!Kf_A!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0a44adc6-7614-4f30-a161-d7b86335ca58_912x912.jpeg 424w, https://substackcdn.com/image/fetch/$s_!Kf_A!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0a44adc6-7614-4f30-a161-d7b86335ca58_912x912.jpeg 848w, https://substackcdn.com/image/fetch/$s_!Kf_A!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0a44adc6-7614-4f30-a161-d7b86335ca58_912x912.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!Kf_A!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0a44adc6-7614-4f30-a161-d7b86335ca58_912x912.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!Kf_A!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0a44adc6-7614-4f30-a161-d7b86335ca58_912x912.jpeg" width="912" height="912" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/0a44adc6-7614-4f30-a161-d7b86335ca58_912x912.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:912,&quot;width&quot;:912,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:300501,&quot;alt&quot;:&quot;Markus Lingman portr&#228;tt&quot;,&quot;title&quot;:null,&quot;type&quot;:&quot;image/jpeg&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="Markus Lingman portr&#228;tt" title="Markus Lingman portr&#228;tt" srcset="https://substackcdn.com/image/fetch/$s_!Kf_A!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0a44adc6-7614-4f30-a161-d7b86335ca58_912x912.jpeg 424w, https://substackcdn.com/image/fetch/$s_!Kf_A!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0a44adc6-7614-4f30-a161-d7b86335ca58_912x912.jpeg 848w, https://substackcdn.com/image/fetch/$s_!Kf_A!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0a44adc6-7614-4f30-a161-d7b86335ca58_912x912.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!Kf_A!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0a44adc6-7614-4f30-a161-d7b86335ca58_912x912.jpeg 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image"><svg role="img" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg"><g><title></title><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a><figcaption class="image-caption">Markus Lingman (foto: Ciprian Gorga)</figcaption></figure></div><p><em>Markus Lingman &#228;r en l&#228;kare och forskare som under sin karri&#228;r r&#246;rt sig mellan klinisk verklighet, ledarskap, akademisk forskning och AI. Som specialist i kardiologi och internmedicin har han sett v&#229;rdens utmaningar p&#229; n&#228;ra h&#229;ll, men ocks&#229; m&#246;jligheterna och v&#228;rdet av den stora m&#228;ngden data som v&#229;rden genererar. I den h&#228;r intervjun ber&#228;ttar han om sin resa fr&#229;n studietiden till chefsuppdrag inom Region Halland och hur intresset f&#246;r AI och informationsdriven v&#229;rd v&#228;xte fram. Vi f&#229;r ta del av konkreta exempel p&#229; hur dataanalys och maskininl&#228;rning kan f&#246;rb&#228;ttra v&#229;rdkedjorna och &#246;ka tryggheten f&#246;r patienterna &#8211; utan att tumma p&#229; integritet och etiska principer. Samtidigt lyfts vikten av samarbete och gemensamma standarder som en avg&#246;rande faktor f&#246;r att Sverige ska lyckas realisera potentialen i informationsdriven v&#229;rd.</em></p><h4><strong>Kan du ber&#228;tta om din bakgrund? Vad ledde in dig p&#229; informationsdriven v&#229;rd och AI?</strong></h4><p>- Stor del av min yrkesbana har varit klinisk som specialist i kardiologi och internmedicin, men f&#246;r knappt 15 &#229;r sedan fick jag mitt f&#246;rsta chefsuppdrag. Jag blev chef f&#246;r l&#228;karna p&#229; en ganska stor sjukhusklinik, men kort d&#228;refter f&#246;ljde ett uppdrag som verksamhetschef och snart ocks&#229; som omr&#229;deschef (det som p&#229; vissa h&#229;ll kallas divisionschef). Sedan n&#229;gra &#229;r arbetar jag i huvudsak i sjukv&#229;rdsledning med strategiska fr&#229;gor. Parallellt med kliniken och chefsbanan har jag bedrivit forskning. Jag disputerade vid Sahlgrenska akademin 2013 d&#228;r jag fortfarande &#228;r affilierad, men har samtidigt en adjungerad professur vid H&#246;gskolan i Halmstad.</p><p>- Tanken om en mer informationsdriven v&#229;rd v&#228;xte fram som ett resultat av mina tre &#8221;ben&#8221; &#8211; klinik, ledning och forskning. Under b&#246;rjan av 2000-talet digitaliserades v&#229;rden. Som en bieffekt b&#246;rjade stora m&#228;ngder data genereras och sparas. M&#228;rkligt nog anv&#228;ndes inte denna guldgruva till insikter i s&#229; stor utstr&#228;ckning. Flertalet siffror som anv&#228;ndes f&#246;r beslut fick plats i ett Excel-ark och det var fortfarande de fyra r&#228;knes&#228;tten fr&#229;n grundskolan som dominerade. Jag bar p&#229; en frustration att v&#229;rden, ett stort komplext adaptivt system, inte hade b&#228;ttre f&#246;ruts&#228;ttningar att fatta r&#228;tt beslut om verksamhetsutveckling. Begreppet &#8221;uppf&#246;ljning&#8221; var vanligt f&#246;rekommande, men uppf&#246;ljning handlar om att titta bak&#229;t. Vid utveckling beh&#246;ver man titta fram&#229;t och kunna beakta m&#228;ngder av faktorer samtidigt s&#229; att det inte dyker upp indirekta negativa bieffekter av de beslut man tar. </p><p>- Jag s&#229;g ett stort behov av att se bortom organisationsgr&#228;nserna och fokusera p&#229; hur det g&#229;r f&#246;r patienterna i relation till allt det vi tillsammans i v&#229;rdens olika delar g&#246;r f&#246;r dem. V&#229;rden &#228;r v&#228;ldigt bra p&#229; att optimera silos och stupr&#246;r &#228;ven om det sker till priset av f&#246;rs&#228;mring n&#229;gon annan stans i systemet. I huvudsak f&#246;r att man inte har faktabaserad &#246;verblick &#246;ver helheten. Informationsdriven v&#229;rd handlar helt enkelt om att arbeta mer faktabaserat, holistiskt och fram&#229;tblickande (prediktivt) genom att nyttigg&#246;ra all den data som vi i v&#229;rden genererar i stora m&#228;ngder varje dag. Bland annat s&#229; utvecklade vi en resursbeskrivningsmodell, Patient Encounter Costing (PEC), f&#246;r att kunna beskriva resurs&#229;tg&#229;ng vid handl&#228;ggning av patienter l&#228;ngs kompletta v&#229;rdkedjor. Motsvarande anv&#228;nder vi nu &#228;ven f&#246;r att r&#228;kna p&#229; klimatp&#229;verkan. Begreppet &#8221;informationsdriven&#8221; hittade jag bara p&#229; f&#246;r att kunna konceptualisera och f&#246;rklara. Id&#233;n kom fr&#229;n <em>The Institute for Information driven medicine </em>p&#229; Harvard.</p><p>- I arbetet med att nyttigg&#246;ra data i v&#229;rden d&#246;k det snabbt upp ett nytt problem &#8211; den var lika komplex och omfattande som v&#229;rden den skall beskriva. Traditionell statistik och &#8221;business intelligence&#8221; r&#228;ckte inte l&#228;ngre till. Det var f&#246;r m&#229;nga faktorer att beakta och samband var inte linj&#228;ra. Samtidigt hade vi turen att maskininl&#228;rning blev tillg&#228;nglig f&#246;r en lite bredare publik, men t&#228;mligen ok&#228;nt inom v&#229;rden. Vi b&#246;rjade samarbeta med forskare inom &#228;mnet p&#229; <em>Harvard Medical School</em> och sedermera <em>H&#246;gskolan i Halmstad</em> och tillsammans f&#246;r&#228;dlade vi data till information som ledde till nya insikter som vi kunde b&#246;rja agera p&#229; och monitorera effekterna av p&#229; ett patientcentriskt s&#228;tt &#8211; essensen i informationsdriven v&#229;rd. Med tiden blev &#228;ven neurala n&#228;tverk till&#228;mpbara och nu utforskar vi m&#246;jligheterna att anv&#228;nda spr&#229;kmodeller.</p><p>- Tack vare de f&#246;rm&#229;gor vi byggde har v&#229;r organisation idag helt andra f&#246;ruts&#228;ttningar att utveckla v&#229;rden och optimera v&#229;rdkedjor faktabaserat &#228;n vad vi hade f&#246;r 10-15 &#229;r sedan.</p><h4><strong>Du b&#246;rjade en civilingenj&#246;rsutbildning men bytte till medicin. Vad har du tagit med dig fr&#229;n ingenj&#246;rsutbildningen?</strong></h4><p>- Innan jag l&#228;ste medicin studerade jag matematik, fysik och programmering, men &#228;ven national-, f&#246;retags- och koncernekonomi samt spr&#229;k inom ramen f&#246;r utbildningen Industriell Ekonomi &#8211; Internationell. Jag fick f&#246;rm&#229;nen att arbeta p&#229; svenska koncerners europakontor under kortare perioder. Att begripa lite mer avancerad matematik har f&#246;rst&#229;s varit stor hj&#228;lp vid AI-till&#228;mpning och forskning. Att kunna ekonomi och ekonometri har varit viktigt eftersom resurshantering &#228;r s&#229; central i v&#229;rden. Inte minst ur ett etiskt perspektiv i prioriteringsarbetet d&#228;r de resurser som st&#229;r till buds m&#229;ste anv&#228;ndas p&#229; det s&#228;tt som skapar st&#246;rst nytta. H&#228;r har v&#229;rden mycket kvar att g&#246;ra. Det &#228;r viktigt att komma ih&#229;g att pengar bara &#228;r ett av m&#229;nga resursslag i v&#229;rden och inte det mest begr&#228;nsande.</p><h4><strong>Kan du dela med dig av n&#229;got exempel d&#228;r ert arbete med informationsdriven v&#229;rd har f&#246;rb&#228;ttrat f&#246;r patienterna i Region Halland?</strong></h4><p>- Halland har idag f&#228;rre v&#229;rdplatser &#228;n andra. Minskningen i v&#229;rdplatser &#228;r i allt v&#228;sentligt en effekt av otillr&#228;cklig kompetenstillg&#229;ng. Minskningen hade varit mycket mer problematiskt om vi inte samtidigt hade kunnat arbeta med hela v&#229;rdkedjorna s&#229; att behovet av inl&#228;ggning p&#229; sjukhus minskade. Bland annat s&#229; anv&#228;nde vi AI f&#246;r att, p&#229; riktigt, f&#246;rst&#229; vad som g&#246;r att en person hamnar p&#229; akutmottagning. Det &#228;r v&#228;ldigt m&#229;nga faktorer som visar sig samverka och d&#228;r vissa kan p&#229;verkas. Med faktabaserad &#246;verblick och prediktionsmodeller har vi kunnat fokusera p&#229; r&#228;tt &#229;tg&#228;rder. Effekten &#228;r att vi inte har fler &#229;terinl&#228;ggningar eller h&#246;gre bel&#228;ggningsniv&#229;er &#228;n tidigare samtidigt som att det g&#229;r &#228;nnu b&#228;ttre f&#246;r patienterna. Vi har ocks&#229; kunnat bygga b&#228;ttre v&#229;rdkedjor f&#246;r enskilda patientgrupper. Personer med hj&#228;rtsvikt i Halland beh&#246;ver betydligt mindre slutenv&#229;rd &#228;n vad som tidigare var fallet. Att arbeta systematiskt grundat i data under l&#229;ng tid har bidragit till att Halland ofta faller v&#228;l ut i nationella kvalitetsj&#228;mf&#246;relser.</p><div class="pullquote"><p>&#8220;Att hitta nya s&#228;tt att bedriva v&#229;rd p&#229; &#228;r helt avg&#246;rande i ljuset av den demografi vi har att se fram emot. &#214;kad precision och prediktion kommer att vara viktiga faktorer och dessa kr&#228;ver data. Jag &#228;r ocks&#229; &#246;vertygad om att m&#229;nga vill f&#229; del av de f&#246;rm&#229;gor som skapas i den stund man sj&#228;lv har blivit eller riskerar att bli sjuk.&#8221;</p></div><h4><strong>Hur ser du p&#229; balansen mellan informationsdriven v&#229;rd och att skydda patienternas integritet? Hur har vi hanterat den balansen i Sverige?</strong></h4><p>- Vi har hanterat k&#228;nslig information i v&#229;rden i alla tider och man kan imponeras av hur s&#228;llan den anv&#228;nds p&#229; otillb&#246;rliga s&#228;tt. H&#228;lso- och sjukv&#229;rdslagen s&#228;ger att vi skall bedriva god v&#229;rd och det inneb&#228;r att den skall vara s&#228;ker, effektiv, kunskapsbaserad, tillg&#228;nglig, individanpassad och j&#228;mlik. Inget av detta skulle kunna garanteras utan faktabaserad utveckling, uppf&#246;ljning och framtidsanalys. D&#228;rf&#246;r ger lagstiftaren v&#229;rdgivaren stora m&#246;jligheter till just detta. Utmaningen &#228;r snarast att den skyddande lagstiftningen kopplas till organisationsstruktur snarare &#228;n till patient (som f&#229;r v&#229;rd i m&#229;nga delar av den fragmenterade organisationen). Kraven p&#229; f&#246;ljsamhet till integritetsskyddande lagstiftning &#228;r dock h&#246;g och medf&#246;r ett antal kontrollmekanismer och begr&#228;nsningar. Ett aktuellt problem &#228;r att mycket lagstiftning har kommit till i en annan tid och d&#228;rmed inte v&#228;gleder i de fr&#229;gor som nu &#228;r aktuella. Det &#228;r tydligt att lagstiftaren, b&#229;de nationellt och p&#229; EU-niv&#229;, vill se mer nyttigg&#246;rande av data f&#246;r att uppn&#229; m&#229;len med v&#229;rden och st&#246;dja forskning. Utvecklingen av lagstiftningen och dess tolkning &#228;r intensiv, men tar tid. </p><p>- Min uppfattning &#228;r att Sverige har varit ganska f&#246;rsiktigt just f&#246;r att skydda integriteten. Jag &#228;r dock lika &#246;vertygad om att man kan skydda integritet och nyttigg&#246;ra data samtidigt. Men man m&#229;ste veta vad man g&#246;r och var gr&#228;nserna g&#229;r. Vinsterna f&#246;r individen och samh&#228;llet &#228;r v&#228;ldigt stora. Att hitta nya s&#228;tt att bedriva v&#229;rd p&#229; &#228;r helt avg&#246;rande i ljuset av den demografi vi har att se fram emot. &#214;kad precision och prediktion kommer att vara viktiga faktorer och dessa kr&#228;ver data. Jag &#228;r ocks&#229; &#246;vertygad om att m&#229;nga vill f&#229; del av de f&#246;rm&#229;gor som skapas i den stund man sj&#228;lv har blivit eller riskerar att bli sjuk. Men f&#246;r att bibeh&#229;lla den tillit som inv&#229;narna har till v&#229;rden idag skall data hanteras med omsorg och i linje med lagstiftarens intentioner.</p><h4><strong>&#197;r 2020 uts&#229;gs du till &#229;rets AI-svensk, bland annat f&#246;r din f&#246;rm&#229;ga "att g&#246;ra det komplexa f&#246;rst&#229;eligt". Tenderar vi i h&#228;lsofr&#229;gor att f&#246;renkla verklighetsbeskrivningar alltf&#246;r mycket?</strong></h4><p>- Absolut. M&#229;nga s&#246;ker enkla svar p&#229; otroligt komplexa fr&#229;gor och snabba l&#246;sningar p&#229; sv&#229;ra problem. Ofta vidtar man &#229;tg&#228;rder i b&#228;sta v&#228;lmening utan att ha koll p&#229; de indirekta negativa konsekvenserna. Detta &#228;r ett klassiskt problem n&#228;r det g&#228;ller styrning som baseras p&#229; indikatorer. Det leder ocks&#229; till att man l&#246;ser problem lokalt &#228;ven om det sker till priset av nya problem n&#229;gon annan stans i systemet eller vid senare tidpunkt. Med &#246;kad faktatillg&#229;ng och en kultur d&#228;r man kan prata om saker s&#229;som det faktiskt &#228;r f&#246;rb&#228;ttras f&#246;ruts&#228;ttningar att attackera problemen p&#229; r&#228;tt s&#228;tt.</p><h4><strong>Du har jobbat med amerikanska samarbetspartners. Vilka &#228;r de st&#246;rsta skillnaderna att arbeta med informationsdriven v&#229;rd och AI i Sverige j&#228;mf&#246;rt med USA?</strong></h4><p>- Det svenska och amerikanska sjukv&#229;rdssystemen &#228;r v&#228;sensskilda. M&#246;jligheterna att anv&#228;nda data hos v&#229;rdgivaren &#228;r st&#246;rre i det amerikanska systemet, men samtidigt pr&#228;glas den till del av att den &#228;r kopplad till ers&#228;ttningar. I USA &#228;r v&#229;rd en industri likt andra industrier som skall agera p&#229; en marknad. Det driver fram vissa typer av AI-l&#246;sningar. Samtidigt har USA en helt enast&#229;ende forskningsmilj&#246; p&#229; ett antal toppuniversitet och &#246;ar av klinisk excellens. Dessa kan vi l&#228;ra av och samarbeta med. Men sedan m&#229;ste till&#228;mpningarna ligga i linje med svenska prioriteringar och s&#228;tt att bedriva v&#229;rd p&#229;. Jag kan avundas hur en del framst&#229;ende akademiska v&#229;rdorganisationer i USA har kunnat integrera forskning och v&#229;rdutf&#246;rande p&#229; s&#228;tt som inte &#228;r genomf&#246;rbara i Sverige. </p><p>- Samtidigt finns det m&#246;jligg&#246;rande f&#246;rdelar i Sverige. Inte minst inv&#229;narnas tillit och j&#228;mlika tillg&#229;ng till v&#229;rden. Det vi har gemensamt &#228;r att beslut alltid blir b&#228;ttre med en b&#228;ttre faktabas. I grunden har vi ocks&#229; samma utmaningar med &#246;kade v&#229;rdbehov i samh&#228;llet och begr&#228;nsad tillg&#229;ng till kompetens &#228;ven om USA demografiska profil &#228;r mer gynnsam &#228;n Sveriges.</p><h4><strong>Hur kan Sverige bli b&#228;ttre p&#229; informationsdriven v&#229;rd och att skapa v&#228;rde av data? Vad &#228;r de viktigaste &#229;tg&#228;rderna?</strong></h4><p>- Det f&#246;rsta korta svaret &#228;r nog &#246;kat samarbete. B&#229;de &#246;ver organisations- och kunskapsgr&#228;nser. Idag sitter m&#229;nga i egna h&#246;rn och f&#246;rs&#246;ker l&#246;sa samma problem och tolka samma lagar. Vi m&#229;ste komma &#246;verens i st&#246;rre utstr&#228;ckning. Det kan handla om standarder, lagtolkningar, vad som &#228;r &#8221;good enough&#8221;. Samtidigt har staten en l&#228;xa att g&#246;ra d&#229; de statliga myndigheter som v&#229;rdutf&#246;raren har att f&#246;rh&#229;lla sig till arbetar alltf&#246;r isolerat fr&#229;n varandra. V&#229;rdgivarna skulle vara betj&#228;nta av &#246;kad v&#228;gledning. </p><p>- Det andra korta svaret &#228;r att inte lockas att se detta som en IT-fr&#229;ga. Att bedriva patientcentrisk, holistisk, fram&#229;tsyftande och faktabaserad v&#229;rd &#228;r en verksamhetsfr&#229;ga d&#228;r kliniker, chefer, IT-specialister, jurister, (h&#228;lso)ekonomer med flera m&#229;ste vara involverade p&#229; ett integrerat s&#228;tt.</p><h4><strong>Hur h&#229;ller du dig sj&#228;lv uppdaterad inom s&#229; snabbt f&#246;r&#228;nderliga omr&#229;den som AI och medicin?</strong></h4><p>- Jag har valt ut ett 20-tal tyngre vetenskapliga tidskrifter inom medicin, AI, patients&#228;kerhet, v&#229;rdkvalitet och h&#228;lsoekonomi som jag bevakar och l&#228;ser l&#246;pande. Det blir m&#229;nga artiklar att g&#229; igenom. Samtidigt har jag lyxen att ha likasinnade och skarpa kollegor runtom i och utanf&#246;r landet som hj&#228;lps &#229;t att bevaka. Det &#228;r v&#228;ldigt stimulerande. N&#228;r det g&#228;ller etik och juridik tycker jag att WHO, EU och OECD med underorgan regelbundet kommer med fina sammanst&#228;llningar och v&#228;gledningar. Sedan l&#228;ser jag f&#246;rst&#229;s en del popul&#228;rmedia ocks&#229;, men den ligger ofta f&#246;r l&#229;ngt efter. Transformermodeller kom 2017, men fick uppm&#228;rksamhet i media f&#246;rst 2024. D&#229; hade allt viktigt redan h&#228;nt.</p><h4><strong>Vad ser du som de mest intressanta AI-trenderna inom v&#229;rden under de kommande 3-5 &#229;ren?</strong></h4><p>- Det jag &#228;r mest nyfiken p&#229; &#228;r hur vi kan nyttigg&#246;ra den enorma m&#228;ngd ostrukturerade data som finns i v&#229;rden, dvs fritext. Den inneh&#229;ller mycket mer information &#228;n den strukturerade och m&#246;jligheterna att skapa v&#228;rden ur den har &#246;kat dramatiskt tack vara transformermodellerna (tex spr&#229;kmodellerna). Det finns dock ett antal oklarheter som vi beh&#246;ver reda i innan vi ser de stora effekterna d&#228;r v&#229;rdpersonal kan avlastas i stor skala.</p><p>- Det jag hoppas p&#229; &#228;r att vi skall kunna till&#228;mpa kliniska beslutsst&#246;d i v&#229;rdvardagen f&#246;r att &#246;ka precision i diagnostik, v&#228;lja skr&#228;ddarsydd behandling fortare f&#246;r den enskilde individen och agera p&#229; risk ist&#228;llet f&#246;r det som redan intr&#228;ffat. F&#246;r detta kr&#228;vs dock f&#246;rtydliganden avseende ansvarsutkr&#228;vande och vilka krav vi skall st&#228;lla p&#229; beslutsst&#246;den. R&#228;cker det att de &#228;r lika bra som en genomsnittlig l&#228;kare? Vilka krav skall vi ha p&#229; transparens och mitigering av bias? Den kommande lagstiftningen uttalar sig om detta, men den beh&#246;ver oms&#228;ttas i praktik och diskuteras i samh&#228;llet. Vi beh&#246;ver f&#229; till en mer gemensam syn p&#229; balansen mellan risk och nytta n&#228;r det g&#228;ller AI precis som vi redan har n&#228;r det g&#228;ller tex l&#228;kemedel.</p><h4><strong>Om du kunde ge ett r&#229;d till unga l&#228;kare och forskare som vill arbeta med AI, vad skulle det vara?</strong></h4><p>- Jag tror att fr&#229;gan &#228;r lite felst&#228;lld. En l&#228;kare eller forskare som inte har hygglig koll p&#229; AI kommer f&#246;rr eller senare bli irrelevant eller &#229;tminstone obsolet. Det arbetas just nu med hur AI-kunskap skall bli en del av all svensk l&#228;karutbildning. Det &#228;r bara lite synd att det skulle dr&#246;ja till 2024. Man beh&#246;ver inte f&#246;rst&#229; detaljerna, men man m&#229;ste ha hygglig k&#228;nnedom om risker och begr&#228;nsningar precis som n&#228;r det g&#228;ller andra verktyg vi anv&#228;nder f&#246;r att hj&#228;lpa patienter.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.informerad.se/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Tack f&#246;r att du l&#228;ser! Prenumerera f&#246;r att f&#229; nya inl&#228;gg och st&#246;dja mitt arbete.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><p><em>Obs! Inneh&#229;llet p&#229; denna blogg &#228;r mina personliga &#229;sikter och representerar inte min arbetsgivare. Jag som utgivare ansvarar inte f&#246;r kommentarsf&#228;ltet. Den som kommenterar ansvarar sj&#228;lv f&#246;r det som skrivs.</em></p>]]></content:encoded></item><item><title><![CDATA[Så landar du drömjobbet inom AI och data: Tips för ansökan och intervju]]></title><description><![CDATA[Att f&#229; dr&#246;mjobbet b&#246;rjar med en v&#228;lplanerad ans&#246;kan och en autentisk intervju.]]></description><link>https://www.informerad.se/p/sa-landar-du-dromjobbet-inom-ai-och</link><guid isPermaLink="false">https://www.informerad.se/p/sa-landar-du-dromjobbet-inom-ai-och</guid><dc:creator><![CDATA[Johan Åhlén]]></dc:creator><pubDate>Fri, 03 Jan 2025 09:20:49 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!QDtb!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F761bb6f8-b43c-4e3d-826a-cabd66e78044_1792x1024.webp" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!QDtb!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F761bb6f8-b43c-4e3d-826a-cabd66e78044_1792x1024.webp" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!QDtb!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F761bb6f8-b43c-4e3d-826a-cabd66e78044_1792x1024.webp 424w, https://substackcdn.com/image/fetch/$s_!QDtb!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F761bb6f8-b43c-4e3d-826a-cabd66e78044_1792x1024.webp 848w, https://substackcdn.com/image/fetch/$s_!QDtb!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F761bb6f8-b43c-4e3d-826a-cabd66e78044_1792x1024.webp 1272w, https://substackcdn.com/image/fetch/$s_!QDtb!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F761bb6f8-b43c-4e3d-826a-cabd66e78044_1792x1024.webp 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!QDtb!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F761bb6f8-b43c-4e3d-826a-cabd66e78044_1792x1024.webp" width="1456" height="832" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/761bb6f8-b43c-4e3d-826a-cabd66e78044_1792x1024.webp&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:832,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:196324,&quot;alt&quot;:&quot;Jobbintervju i kontorsmilj&#246;&quot;,&quot;title&quot;:null,&quot;type&quot;:&quot;image/webp&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="Jobbintervju i kontorsmilj&#246;" title="Jobbintervju i kontorsmilj&#246;" srcset="https://substackcdn.com/image/fetch/$s_!QDtb!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F761bb6f8-b43c-4e3d-826a-cabd66e78044_1792x1024.webp 424w, https://substackcdn.com/image/fetch/$s_!QDtb!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F761bb6f8-b43c-4e3d-826a-cabd66e78044_1792x1024.webp 848w, https://substackcdn.com/image/fetch/$s_!QDtb!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F761bb6f8-b43c-4e3d-826a-cabd66e78044_1792x1024.webp 1272w, https://substackcdn.com/image/fetch/$s_!QDtb!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F761bb6f8-b43c-4e3d-826a-cabd66e78044_1792x1024.webp 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image"><svg role="img" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg"><g><title></title><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>Att arbeta inom AI, analys och data &#228;r riktigt sp&#228;nnande. Inte minst p&#229; Region Sk&#229;ne, som g&#229;r i frontlinjen genom regional <a href="https://www.skane.se/om-region-skane/nyheter/2024/regionstyrelsen-tog-beslut-om-strategier-for-framtiden/">AI-strategi</a> och stora satsningar inom omr&#229;det. H&#228;lso- och sjukv&#229;rd &#228;r ett av de anv&#228;ndningsomr&#229;den som har st&#246;rst potential att dra nytta av AI och modern teknik.</p><p>Konkurrensen om dessa roller &#228;r ofta h&#246;g. Efter att ha genomf&#246;rt m&#228;ngder av intervjuer senaste &#229;ren och pl&#246;jt igenom ett or&#228;kneligt antal CV:n har jag h&#228;r samlat mina b&#228;sta tips f&#246;r hur du lyckas med din ans&#246;kan och intervju!</p><h3>Tips f&#246;r ditt CV / ans&#246;kan</h3><p>Din ans&#246;kan och ditt CV &#228;r din intr&#228;desbiljett f&#246;r att komma p&#229; intervju. Det &#228;r ocks&#229; f&#246;rsta intrycket arbetsgivaren f&#229;r av dig.</p><ul><li><p>Skr&#228;ddarsy din ans&#246;kan. L&#228;s noga igenom annonsen och se till att du har klart f&#246;r dig vad som &#228;r absoluta krav och vad som bara &#228;r meriterande. Fundera ocks&#229; &#246;ver vilka andra styrkor och erfarenheter du kan ha som &#228;r anv&#228;ndbara f&#246;r jobbet. </p></li><li><p>Var personlig. Var tydlig med varf&#246;r du s&#246;ker just denna roll och vad du tycker &#228;r viktigt f&#246;r att trivas. Skriv g&#228;rna ocks&#229; om hur du &#228;r som person och vilka intressen och hobbies du har.</p></li><li><p>Var &#228;rlig. Se till att det som st&#229;r i sin ans&#246;kan och ditt CV ocks&#229; st&#228;mmer med exempelvis vad som st&#229;r p&#229; din LinkedIn-profil. Det &#228;r inget konstigt med att lyfta andra saker i ans&#246;kan &#228;n p&#229; LinkedIn, men det m&#229;ste vara korrekt. Var ocks&#229; genuin med hur du beskriver dig sj&#228;lv.</p></li><li><p>Visa g&#228;rna arbetsprover. Har du n&#229;gra artiklar, presentationer, projekt eller annat du kan visa upp? L&#228;nka g&#228;rna till dem i din ans&#246;kan.</p></li><li><p>Var tydlig och strukturerad. H&#229;ll nere l&#228;ngden p&#229; ditt CV och ans&#246;kan, inga romaner allts&#229;. Se bara till att det &#228;r tydligt varf&#246;r du &#228;r en bra kandidat f&#246;r rollen. </p></li><li><p>L&#229;t g&#228;rna en v&#228;n kika p&#229; din ans&#246;kan och ge dig feedback!</p></li></ul><h3>Tips inf&#246;r intervju</h3><p>Intervjun &#228;r din chans att visa vem du &#228;r och varf&#246;r du &#228;r r&#228;tt person f&#246;r jobbet. Det &#228;r inte bara en m&#246;jlighet att bevisa din kompetens utan ocks&#229; att visa p&#229; s&#229;dant som inte framkommer av CV:t.</p><ul><li><p>F&#246;rst&#229; rollen. L&#228;s annonsen noggrant inf&#246;r intervjun och fundera &#246;ver om du sj&#228;lv har n&#229;gra fr&#229;gor. F&#246;rbered dig p&#229; fr&#229;gor du tror kan komma upp. </p></li><li><p>Var dig sj&#228;lv. Det &#228;r helt okej att vara nerv&#246;s inf&#246;r intervjun och det &#228;r fullt m&#228;nskligt. Var trygg i dina styrkor och var &#246;ppen med vad du vill utveckla. Var ocks&#229; &#246;ppen med vad som &#228;r viktigt f&#246;r just dig. Autenticitet vinner alltid.</p></li><li><p>Svara sakligt p&#229; fr&#229;gorna. Det &#228;r helt okej att t&#228;nka efter en stund f&#246;r att ge ett bra svar. Resonera g&#228;rna, ofta &#228;r det &#246;ppna fr&#229;gor utan ett sj&#228;lvklart svar. Ge g&#228;rna exempel som visar att du har verklig erfarenhet av &#228;mnet. Kommer du inte p&#229; n&#229;got svar s&#229; var bara &#228;rlig om det. Chansa inte genom att &#8220;snacka p&#229;&#8221; och hoppas att n&#229;got av det du s&#228;ger var det &#246;nskade svaret.</p></li><li><p>Ge lagom l&#229;nga svar. Detta &#228;r sv&#229;rt, och n&#228;r vi blir nerv&#246;sa tenderar vi ibland att prata antingen f&#246;r mycket eller f&#246;r lite beroende p&#229; v&#229;r personlighet. Om du &#228;r os&#228;ker, kolla g&#228;rna med intervjuarna om ni h&#229;ller tidsplanen f&#246;r intervjun.</p></li><li><p>Ett s&#228;tt att f&#246;rbereda svar &#228;r STAR-metoden (Situation, Task (Uppgift), Action (Vad du gjorde), Resultat). Fundera &#246;ver ett par relevanta exempel p&#229; situationer d&#228;r du haft en uppgift, vad du gjorde och vad resultatet blev. Det beh&#246;ver inte bara vara arbetssituationer.</p></li></ul><h3>Slutsats</h3><p>Att s&#246;ka jobb &#228;r en process som kr&#228;ver tid och energi. Genom att skr&#228;ddarsy din ans&#246;kan, f&#246;rbereda dig noggrant inf&#246;r intervjuer och visa &#228;kta engagemang kan du &#246;ka dina chanser att landa dr&#246;mjobbet. Och kom ih&#229;g &#8211; varje erfarenhet, oavsett om det leder till en anst&#228;llning eller inte, &#228;r ett steg n&#228;rmare ditt m&#229;l.</p><p>Lycka till!</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.informerad.se/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Tack f&#246;r att du l&#228;ser! Prenumerera f&#246;r att f&#229; nya inl&#228;gg och st&#246;dja mitt arbete.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><p>Obs! Inneh&#229;llet p&#229; denna blogg &#228;r mina personliga &#229;sikter och representerar inte min arbetsgivare. Jag som utgivare ansvarar inte f&#246;r kommentarsf&#228;ltet. Den som kommenterar ansvarar sj&#228;lv f&#246;r det som skrivs.</p>]]></content:encoded></item><item><title><![CDATA[Testa dina AI-kunskaper med detta quiz]]></title><description><![CDATA[Hur mycket av det vi ser i vardagen &#228;r AI? L&#228;r dig mer om vad AI faktiskt inneb&#228;r och hur det skiljer sig fr&#229;n annan teknologi.]]></description><link>https://www.informerad.se/p/testa-dina-ai-kunskaper-med-detta</link><guid isPermaLink="false">https://www.informerad.se/p/testa-dina-ai-kunskaper-med-detta</guid><dc:creator><![CDATA[Johan Åhlén]]></dc:creator><pubDate>Thu, 02 Jan 2025 09:50:25 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!Ez4a!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe418b7d3-62af-4d21-bd30-a6ddf8e60cd6_1792x1024.webp" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!Ez4a!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe418b7d3-62af-4d21-bd30-a6ddf8e60cd6_1792x1024.webp" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!Ez4a!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe418b7d3-62af-4d21-bd30-a6ddf8e60cd6_1792x1024.webp 424w, https://substackcdn.com/image/fetch/$s_!Ez4a!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe418b7d3-62af-4d21-bd30-a6ddf8e60cd6_1792x1024.webp 848w, https://substackcdn.com/image/fetch/$s_!Ez4a!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe418b7d3-62af-4d21-bd30-a6ddf8e60cd6_1792x1024.webp 1272w, https://substackcdn.com/image/fetch/$s_!Ez4a!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe418b7d3-62af-4d21-bd30-a6ddf8e60cd6_1792x1024.webp 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!Ez4a!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe418b7d3-62af-4d21-bd30-a6ddf8e60cd6_1792x1024.webp" width="1456" height="832" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/e418b7d3-62af-4d21-bd30-a6ddf8e60cd6_1792x1024.webp&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:832,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:317776,&quot;alt&quot;:&quot;Robot som svarar p&#229; ett fr&#229;geformul&#228;r&quot;,&quot;title&quot;:null,&quot;type&quot;:&quot;image/webp&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="Robot som svarar p&#229; ett fr&#229;geformul&#228;r" title="Robot som svarar p&#229; ett fr&#229;geformul&#228;r" srcset="https://substackcdn.com/image/fetch/$s_!Ez4a!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe418b7d3-62af-4d21-bd30-a6ddf8e60cd6_1792x1024.webp 424w, https://substackcdn.com/image/fetch/$s_!Ez4a!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe418b7d3-62af-4d21-bd30-a6ddf8e60cd6_1792x1024.webp 848w, https://substackcdn.com/image/fetch/$s_!Ez4a!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe418b7d3-62af-4d21-bd30-a6ddf8e60cd6_1792x1024.webp 1272w, https://substackcdn.com/image/fetch/$s_!Ez4a!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe418b7d3-62af-4d21-bd30-a6ddf8e60cd6_1792x1024.webp 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image"><svg role="img" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg"><g><title></title><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://ai-quiz.informerad.se/&quot;,&quot;text&quot;:&quot;Starta AI-quizet nu!&quot;,&quot;action&quot;:null,&quot;class&quot;:&quot;button-wrapper&quot;}" data-component-name="ButtonCreateButton"><a class="button primary button-wrapper" href="https://ai-quiz.informerad.se/"><span>Starta AI-quizet nu!</span></a></p><p>Att f&#246;rst&#229; vad artificiell intelligens (AI) &#228;r blir allt viktigare i en v&#228;rld d&#228;r teknologin utvecklas i rasande takt. Genom att delta i detta quiz kan du:</p><ul><li><p><strong>F&#229; b&#228;ttre koll p&#229; AI:</strong> L&#228;r dig mer om vad AI faktiskt inneb&#228;r och hur det skiljer sig fr&#229;n annan teknologi. Efter varje svar f&#229;r du en kort f&#246;rklaring om varf&#246;r det &#228;r (eller inte &#228;r) AI.</p></li><li><p><strong>Utmana dina f&#246;rdomar:</strong> Det &#228;r l&#228;tt att tro att AI &#228;r n&#229;got mystiskt och magiskt &#8211; men hur mycket av det vi ser i vardagen &#228;r faktiskt AI? Vissa exempel kan vara lurigare &#228;n du tror!</p></li><li><p><strong>Ha kul:</strong> Oavsett om du &#228;r tekniskt insatt eller bara nyfiken, &#228;r quizet ett roligt s&#228;tt att utforska AI tillsammans med andra.</p></li></ul><p>Det &#228;r superenkelt att komma ig&#229;ng! Klicka p&#229; knappen nedan f&#246;r att starta quizet och dela det sedan med dina v&#228;nner. Se vem som f&#229;r flest r&#228;tt &#8211; och vem som beh&#246;ver l&#228;ra sig mer!</p><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://ai-quiz.informerad.se/&quot;,&quot;text&quot;:&quot;Starta AI-quizet nu!&quot;,&quot;action&quot;:null,&quot;class&quot;:null}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://ai-quiz.informerad.se/"><span>Starta AI-quizet nu!</span></a></p><p>Hoppas att quizet skall vara till n&#246;je! Om du har feedback eller egna fr&#229;gor som du tycker borde l&#228;ggas till, tveka inte att h&#246;ra av dig.</p><p><a href="https://www.informerad.se/p/vad-ar-ai">L&#228;s mer om vad AI &#228;r</a></p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.informerad.se/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Tack f&#246;r att du l&#228;ser! Prenumerera f&#246;r att f&#229; nya inl&#228;gg och st&#246;dja mitt arbete.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><p><em>Obs! Inneh&#229;llet p&#229; denna blogg &#228;r mina personliga &#229;sikter och representerar inte min arbetsgivare. Jag som utgivare ansvarar inte f&#246;r kommentarsf&#228;ltet. Den som kommenterar ansvarar sj&#228;lv f&#246;r det som skrivs.</em></p>]]></content:encoded></item><item><title><![CDATA[Skillnaden mellan data och information: från råa siffror till mening]]></title><description><![CDATA[Genom att f&#246;rst&#229; denna skillnad kan vi bli b&#228;ttre p&#229; att skapa v&#228;rde av v&#229;r data och fatta b&#228;ttre informerade beslut.]]></description><link>https://www.informerad.se/p/skillnaden-mellan-data-och-information</link><guid isPermaLink="false">https://www.informerad.se/p/skillnaden-mellan-data-och-information</guid><dc:creator><![CDATA[Johan Åhlén]]></dc:creator><pubDate>Sun, 20 Oct 2024 13:05:21 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!A22k!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa800da5e-7013-4ed8-8eb5-b643662c4560_1792x1024.webp" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!A22k!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa800da5e-7013-4ed8-8eb5-b643662c4560_1792x1024.webp" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!A22k!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa800da5e-7013-4ed8-8eb5-b643662c4560_1792x1024.webp 424w, https://substackcdn.com/image/fetch/$s_!A22k!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa800da5e-7013-4ed8-8eb5-b643662c4560_1792x1024.webp 848w, https://substackcdn.com/image/fetch/$s_!A22k!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa800da5e-7013-4ed8-8eb5-b643662c4560_1792x1024.webp 1272w, https://substackcdn.com/image/fetch/$s_!A22k!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa800da5e-7013-4ed8-8eb5-b643662c4560_1792x1024.webp 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!A22k!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa800da5e-7013-4ed8-8eb5-b643662c4560_1792x1024.webp" width="1456" height="832" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/a800da5e-7013-4ed8-8eb5-b643662c4560_1792x1024.webp&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:832,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:791676,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/webp&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!A22k!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa800da5e-7013-4ed8-8eb5-b643662c4560_1792x1024.webp 424w, https://substackcdn.com/image/fetch/$s_!A22k!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa800da5e-7013-4ed8-8eb5-b643662c4560_1792x1024.webp 848w, https://substackcdn.com/image/fetch/$s_!A22k!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa800da5e-7013-4ed8-8eb5-b643662c4560_1792x1024.webp 1272w, https://substackcdn.com/image/fetch/$s_!A22k!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa800da5e-7013-4ed8-8eb5-b643662c4560_1792x1024.webp 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image"><svg role="img" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg"><g><title></title><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>Att f&#246;rst&#229; skillnaden mellan <em>data</em> och <em>information</em> kan vara lite av en utmaning. &#196;n v&#228;rre, om du s&#246;ker upp ordet <em>data</em> i en etablerad ordbok, som exempelvis <a href="https://dictionary.cambridge.org/dictionary/english/data">Cambridge Dictionary</a>, s&#229; &#228;r det risk att du blir fortsatt f&#246;rvirrad:</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!GROI!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc8eb1a34-ec48-40a6-9fb1-26850fdab476_748x250.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!GROI!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc8eb1a34-ec48-40a6-9fb1-26850fdab476_748x250.png 424w, https://substackcdn.com/image/fetch/$s_!GROI!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc8eb1a34-ec48-40a6-9fb1-26850fdab476_748x250.png 848w, https://substackcdn.com/image/fetch/$s_!GROI!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc8eb1a34-ec48-40a6-9fb1-26850fdab476_748x250.png 1272w, https://substackcdn.com/image/fetch/$s_!GROI!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc8eb1a34-ec48-40a6-9fb1-26850fdab476_748x250.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!GROI!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc8eb1a34-ec48-40a6-9fb1-26850fdab476_748x250.png" width="748" height="250" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/c8eb1a34-ec48-40a6-9fb1-26850fdab476_748x250.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:250,&quot;width&quot;:748,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:18363,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!GROI!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc8eb1a34-ec48-40a6-9fb1-26850fdab476_748x250.png 424w, https://substackcdn.com/image/fetch/$s_!GROI!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc8eb1a34-ec48-40a6-9fb1-26850fdab476_748x250.png 848w, https://substackcdn.com/image/fetch/$s_!GROI!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc8eb1a34-ec48-40a6-9fb1-26850fdab476_748x250.png 1272w, https://substackcdn.com/image/fetch/$s_!GROI!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc8eb1a34-ec48-40a6-9fb1-26850fdab476_748x250.png 1456w" sizes="100vw"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image"><svg role="img" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg"><g><title></title><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a><figcaption class="image-caption">Cambridge Dictionary: Data</figcaption></figure></div><p>Det finns ingen vedertagen och universellt accepterad definition av begreppen <em>data</em> och <em>information</em>, utan olika experter kommer att ge olika svar. Skillnaden &#228;r dock viktig, bland annat f&#246;r att inte bli lurad att tro att du har information n&#228;r du i sj&#228;lva verket bara har data.</p><p>F&#246;renklat kan skillnaden beskrivas s&#229; h&#228;r:</p><ul><li><p>Data &#228;r en sorts <em>r&#229;vara</em> som representerar information. Data kan lagras och behandlas. Exempelvis en databas kan lagra och behandla data, men data kan ocks&#229; lagras som en fil.</p></li><li><p>Information &#228;r sj&#228;lva inneb&#246;rden eller <em>betydelsen</em> av data. Ett blodtrycksv&#228;rde p&#229; 125/75 &#228;r information, medan sj&#228;lva siffrorna 125 och 75 &#228;r data som inte skulle betyda n&#229;got utan sammanhang. Information i sig kan inte lagras, men data som representerar informationen g&#229;r att lagra. </p></li></ul><p>En faktauppgift som att &#8220;r&#246;str&#228;tts&#229;ldern i Sverige &#228;r 18 &#229;r&#8221; &#228;r information. Om du skulle skriva faktauppgiften i ett dokument och spara som en fil s&#229; har du omvandlat informationen till data.</p><div class="pullquote"><p>Om du inte skiljer mellan data och information s&#229; kan du g&#246;ra misstaget att tro att du har information n&#228;r du egentligen bara har data. </p></div><h3>Ett exempel p&#229; skillnaden</h3><p>L&#229;t oss kika p&#229; ett exempel. Kan du se skillnad p&#229; dessa bilder?</p><div class="captioned-image-container"><figure><a class="image-link image2" target="_blank" href="https://substackcdn.com/image/fetch/$s_!OiuU!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3df4fba2-0319-44dd-939f-c0a34f99c3a3_268x153.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!OiuU!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3df4fba2-0319-44dd-939f-c0a34f99c3a3_268x153.png 424w, https://substackcdn.com/image/fetch/$s_!OiuU!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3df4fba2-0319-44dd-939f-c0a34f99c3a3_268x153.png 848w, https://substackcdn.com/image/fetch/$s_!OiuU!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3df4fba2-0319-44dd-939f-c0a34f99c3a3_268x153.png 1272w, https://substackcdn.com/image/fetch/$s_!OiuU!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3df4fba2-0319-44dd-939f-c0a34f99c3a3_268x153.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!OiuU!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3df4fba2-0319-44dd-939f-c0a34f99c3a3_268x153.png" width="268" height="153" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/3df4fba2-0319-44dd-939f-c0a34f99c3a3_268x153.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:153,&quot;width&quot;:268,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:121552,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!OiuU!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3df4fba2-0319-44dd-939f-c0a34f99c3a3_268x153.png 424w, https://substackcdn.com/image/fetch/$s_!OiuU!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3df4fba2-0319-44dd-939f-c0a34f99c3a3_268x153.png 848w, https://substackcdn.com/image/fetch/$s_!OiuU!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3df4fba2-0319-44dd-939f-c0a34f99c3a3_268x153.png 1272w, https://substackcdn.com/image/fetch/$s_!OiuU!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3df4fba2-0319-44dd-939f-c0a34f99c3a3_268x153.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></a><figcaption class="image-caption">scientist.tif</figcaption></figure></div><div class="captioned-image-container"><figure><a class="image-link image2" target="_blank" href="https://substackcdn.com/image/fetch/$s_!dJYT!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4353c9d8-cc23-44cb-825f-828d8fec843d_268x153.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!dJYT!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4353c9d8-cc23-44cb-825f-828d8fec843d_268x153.png 424w, https://substackcdn.com/image/fetch/$s_!dJYT!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4353c9d8-cc23-44cb-825f-828d8fec843d_268x153.png 848w, https://substackcdn.com/image/fetch/$s_!dJYT!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4353c9d8-cc23-44cb-825f-828d8fec843d_268x153.png 1272w, https://substackcdn.com/image/fetch/$s_!dJYT!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4353c9d8-cc23-44cb-825f-828d8fec843d_268x153.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!dJYT!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4353c9d8-cc23-44cb-825f-828d8fec843d_268x153.png" width="268" height="153" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/4353c9d8-cc23-44cb-825f-828d8fec843d_268x153.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:153,&quot;width&quot;:268,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:121552,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!dJYT!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4353c9d8-cc23-44cb-825f-828d8fec843d_268x153.png 424w, https://substackcdn.com/image/fetch/$s_!dJYT!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4353c9d8-cc23-44cb-825f-828d8fec843d_268x153.png 848w, https://substackcdn.com/image/fetch/$s_!dJYT!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4353c9d8-cc23-44cb-825f-828d8fec843d_268x153.png 1272w, https://substackcdn.com/image/fetch/$s_!dJYT!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4353c9d8-cc23-44cb-825f-828d8fec843d_268x153.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></a><figcaption class="image-caption">scientist.png</figcaption></figure></div><p>Hur du &#228;n j&#228;mf&#246;r s&#229; kommer du inte kunna <em>se</em> n&#229;gon skillnad. Bilderna visar exakt samma sak. De har exakt samma utseende och d&#228;rmed samma information.</p><p>Om du d&#228;remot skulle j&#228;mf&#246;ra bildfilerna, <em>scientist.tif</em> och <em>scientist.png</em>, s&#229; &#228;r de olika b&#229;de till inneh&#229;ll och storlek. Skillnaden beror p&#229; att bilden &#228;r sparad i tv&#229; olika filformat. Ett filformat &#228;r ett slags regelverk f&#246;r hur information (exempelvis en bild) representeras som data. Nedan kan du se exempel p&#229; inneh&#229;llet i filerna:</p><div class="captioned-image-container"><figure><a class="image-link image2" target="_blank" href="https://substackcdn.com/image/fetch/$s_!CAm6!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F485eee05-336e-4980-9134-5fa48fac1b48_676x137.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!CAm6!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F485eee05-336e-4980-9134-5fa48fac1b48_676x137.png 424w, https://substackcdn.com/image/fetch/$s_!CAm6!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F485eee05-336e-4980-9134-5fa48fac1b48_676x137.png 848w, https://substackcdn.com/image/fetch/$s_!CAm6!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F485eee05-336e-4980-9134-5fa48fac1b48_676x137.png 1272w, https://substackcdn.com/image/fetch/$s_!CAm6!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F485eee05-336e-4980-9134-5fa48fac1b48_676x137.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!CAm6!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F485eee05-336e-4980-9134-5fa48fac1b48_676x137.png" width="676" height="137" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/485eee05-336e-4980-9134-5fa48fac1b48_676x137.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:137,&quot;width&quot;:676,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:45147,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!CAm6!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F485eee05-336e-4980-9134-5fa48fac1b48_676x137.png 424w, https://substackcdn.com/image/fetch/$s_!CAm6!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F485eee05-336e-4980-9134-5fa48fac1b48_676x137.png 848w, https://substackcdn.com/image/fetch/$s_!CAm6!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F485eee05-336e-4980-9134-5fa48fac1b48_676x137.png 1272w, https://substackcdn.com/image/fetch/$s_!CAm6!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F485eee05-336e-4980-9134-5fa48fac1b48_676x137.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></a><figcaption class="image-caption">Filinneh&#229;ll i scientist.tif</figcaption></figure></div><div class="captioned-image-container"><figure><a class="image-link image2" target="_blank" href="https://substackcdn.com/image/fetch/$s_!jNJ1!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F061d3803-21df-440e-b1a3-fe760c972aaf_680x136.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!jNJ1!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F061d3803-21df-440e-b1a3-fe760c972aaf_680x136.png 424w, https://substackcdn.com/image/fetch/$s_!jNJ1!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F061d3803-21df-440e-b1a3-fe760c972aaf_680x136.png 848w, https://substackcdn.com/image/fetch/$s_!jNJ1!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F061d3803-21df-440e-b1a3-fe760c972aaf_680x136.png 1272w, https://substackcdn.com/image/fetch/$s_!jNJ1!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F061d3803-21df-440e-b1a3-fe760c972aaf_680x136.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!jNJ1!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F061d3803-21df-440e-b1a3-fe760c972aaf_680x136.png" width="680" height="136" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/061d3803-21df-440e-b1a3-fe760c972aaf_680x136.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:136,&quot;width&quot;:680,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:38833,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!jNJ1!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F061d3803-21df-440e-b1a3-fe760c972aaf_680x136.png 424w, https://substackcdn.com/image/fetch/$s_!jNJ1!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F061d3803-21df-440e-b1a3-fe760c972aaf_680x136.png 848w, https://substackcdn.com/image/fetch/$s_!jNJ1!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F061d3803-21df-440e-b1a3-fe760c972aaf_680x136.png 1272w, https://substackcdn.com/image/fetch/$s_!jNJ1!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F061d3803-21df-440e-b1a3-fe760c972aaf_680x136.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></a><figcaption class="image-caption">Filinneh&#229;ll i scientist.png</figcaption></figure></div><p>F&#246;r en bild s&#229; &#228;r sj&#228;lva utseendet information medan bildfilen som representerar den &#228;r data.</p><p>Sammanfattningsvis s&#229; &#228;r det allts&#229; s&#229; h&#228;r i detta exempel:</p><ul><li><p>Informationen i bilderna (utseendet) &#228;r exakt samma.</p></li><li><p>Data (den digitala representationen) &#228;r olika f&#246;r bilderna eftersom de har olika filformat.</p></li></ul><p>Detta samband mellan information och data g&#228;ller inte bara bilder, utan det &#228;r p&#229; precis samma s&#228;tt vad det &#228;n g&#228;ller f&#246;r information s&#229;som dokument, video, blodtrycksv&#228;rden, osv. &#196;ven de kan representeras i olika format.</p><h3>Ett sifferexempel</h3><p>Om vi v&#228;nder p&#229; det d&#229;, skulle samma data kunna utg&#246;ra olika information?</p><p>G&#246;r tankeexperimentet att du har en serie siffror (v&#229;r data):</p><pre><code>31, 29, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31</code></pre><p>Har du n&#229;gon aning om vad dessa siffror kan betyda?</p><ul><li><p>&#197;ldrar p&#229; personer</p></li><li><p>Elevers po&#228;ngresultat p&#229; ett prov</p></li><li><p>Temperaturm&#228;tningar</p></li><li><p>Etc</p></li></ul><p>Utifr&#229;n enbart siffror (v&#229;r data) s&#229; g&#229;r det inte att veta helt s&#228;kert vad de beskriver (informationen). Utan sammanhang &#228;r data inte tolkningsbart.</p><p>Just i det h&#228;r fallet finns faktiskt &#228;nd&#229; en kvalificerad gissning. Du kanske redan har kommit p&#229; det, men om du t&#228;nker m&#229;nader och dagar under ett skott&#229;r s&#229; kommer du landa i precis denna sifferserie. Det &#228;r dessutom en h&#246;gst kvalificerad gissning eftersom det vore ett riktigt sammantr&#228;ffande att lyckas f&#229; denna sifferserie p&#229; annat s&#228;tt &#228;n genom att r&#228;kna antalet dagar per m&#229;nad. </p><p>Det &#228;r &#228;nd&#229; en tolkningsfr&#229;ga vad siffrorna betyder. Utan sammanhang g&#229;r det inte att veta helt s&#228;kert vilken tolkning som &#228;r r&#228;tt. Data i sig har allts&#229; ingen betydelse, utan det &#228;r f&#246;rst n&#228;r den s&#228;tts i ett sammanhang som den kan omvandlas till information.</p><h3>Interoperabilitet och standarder</h3><p>Att samma information finns i olika system inneb&#228;r inte n&#246;dv&#228;ndigtvis att de kan dela informationen mellan varandra. Eftersom data kan tolkas p&#229; s&#229; m&#229;nga olika s&#228;tt s&#229; beh&#246;vs standardiserade regelverk och tolkningar f&#246;r att uppn&#229; <a href="https://interoperabilitet.se/">interoperabilitet</a>.</p><p>Det finns m&#228;ngder av s&#229;dana h&#228;r standarder. Ett exempel &#228;r de bildfilformat som finns. De beskriver helt enkelt hur en bild p&#229; ett standardiserat s&#228;tt omvandlas till data. Tack vare dessa standardiserade bildfilformat, och standarden HTML, f&#229;r vi interoperabilitet mellan olika webbl&#228;sare s&#229; att det ser likadant ut oavsett vilken webbl&#228;sare du v&#228;ljer.</p><p>Det finns ocks&#229; standarder av andra slag som beskriver hur information representeras som data. <a href="https://www.hl7.org/fhir/">FHIR</a> och <a href="https://openehr.org/">openEHR</a> &#228;r internationella och &#246;ppna standarder f&#246;r datautbyte inom h&#228;lso- och sjukv&#229;rd som specificerar just hur h&#228;lsoinformation skall beskrivas som data. Detta &#228;r &#229;terigen ett exempel p&#229; interoperabilitet, att m&#246;jligg&#246;ra f&#246;r olika journalsystem, laboratorieutrustning och andra applikationer att dela information med varandra. </p><p>Det finns fler faktorer f&#246;r att uppn&#229; interoperabilitet, men ett grundl&#228;ggande krav &#228;r att informationen representeras som data p&#229; ett standardiserat s&#228;tt. </p><h3>Varf&#246;r &#228;r skillnaden viktig?</h3><p>T&#228;nk dig att du skulle komma &#229;t en <em>databas</em>. Det du kommer hitta i databasen &#228;r data men det kommer att vara mycket sv&#229;rt att f&#246;rst&#229; betydelsen av den om du inte har n&#229;gon beskrivning av sammanhanget och hur databasen &#228;r strukturerad. Du har allts&#229; data men inte information.</p><p>Detsamma g&#228;ller API:er. Om du inte har n&#229;gon beskrivning eller sammanhanget framg&#229;r s&#229; kommer data fr&#229;n API:erna l&#228;tt att misstolkas.</p><p>Se allts&#229; till att du vet hur du skall tolka din data s&#229; att den blir information. Om du inte skiljer mellan data och information s&#229; kan du g&#246;ra misstaget att tro att du har information n&#228;r du egentligen bara har data. </p><h3>Slutsats</h3><p>Sammanfattningsvis s&#229; &#228;r f&#246;rst&#229;else f&#246;r skillnaden mellan data och information grundl&#228;ggande f&#246;r att kunna f&#229; fram bra beslutsunderlag. Data &#228;r det r&#229;material vi samlar in, medan information &#228;r den tolkade och bearbetade versionen som ger oss insikter och sammanhang. Genom att f&#246;rst&#229; denna skillnad kan vi bli b&#228;ttre p&#229; att skapa v&#228;rde av v&#229;r data och fatta b&#228;ttre informerade beslut.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.informerad.se/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Tack f&#246;r att du l&#228;ser. Prenumerera h&#228;r!</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><p><em>Obs! Inneh&#229;llet p&#229; denna blogg &#228;r mina personliga &#229;sikter och representerar inte min arbetsgivare. Jag som utgivare ansvarar inte f&#246;r kommentarsf&#228;ltet. Den som kommenterar ansvarar sj&#228;lv f&#246;r det som skrivs.</em></p>]]></content:encoded></item><item><title><![CDATA[Stefan Jovinge, a pioneer's journey in cardiac intensive care and AI]]></title><description><![CDATA[From leading institutions in the USA, Research Director Stefan Jovinge shares his insights on how experiences in AI and health data have shaped his vision for the future of Life Science in Sweden.]]></description><link>https://www.informerad.se/p/stefan-jovinge-a-pioneers-journey</link><guid isPermaLink="false">https://www.informerad.se/p/stefan-jovinge-a-pioneers-journey</guid><dc:creator><![CDATA[Johan Åhlén]]></dc:creator><pubDate>Mon, 16 Sep 2024 19:05:38 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!AM8k!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F435eff88-2622-411c-9bec-7df6e1d7d426_827x1079.jpeg" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!AM8k!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F435eff88-2622-411c-9bec-7df6e1d7d426_827x1079.jpeg" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!AM8k!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F435eff88-2622-411c-9bec-7df6e1d7d426_827x1079.jpeg 424w, https://substackcdn.com/image/fetch/$s_!AM8k!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F435eff88-2622-411c-9bec-7df6e1d7d426_827x1079.jpeg 848w, https://substackcdn.com/image/fetch/$s_!AM8k!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F435eff88-2622-411c-9bec-7df6e1d7d426_827x1079.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!AM8k!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F435eff88-2622-411c-9bec-7df6e1d7d426_827x1079.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!AM8k!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F435eff88-2622-411c-9bec-7df6e1d7d426_827x1079.jpeg" width="827" height="1079" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/435eff88-2622-411c-9bec-7df6e1d7d426_827x1079.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1079,&quot;width&quot;:827,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:529792,&quot;alt&quot;:&quot;Stefan Jovinge&quot;,&quot;title&quot;:null,&quot;type&quot;:&quot;image/jpeg&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="Stefan Jovinge" title="Stefan Jovinge" srcset="https://substackcdn.com/image/fetch/$s_!AM8k!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F435eff88-2622-411c-9bec-7df6e1d7d426_827x1079.jpeg 424w, https://substackcdn.com/image/fetch/$s_!AM8k!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F435eff88-2622-411c-9bec-7df6e1d7d426_827x1079.jpeg 848w, https://substackcdn.com/image/fetch/$s_!AM8k!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F435eff88-2622-411c-9bec-7df6e1d7d426_827x1079.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!AM8k!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F435eff88-2622-411c-9bec-7df6e1d7d426_827x1079.jpeg 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image"><svg role="img" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg"><g><title></title><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a><figcaption class="image-caption">Stefan Jovinge</figcaption></figure></div><p><em>In this interview, we meet Stefan Jovinge, Research Director at Sk&#229;ne University Hospital, who has made an impressive journey from the Karolinska Institute to leading institutions in the USA, including Stanford and Harvard. With a unique combination of clinical expertise and advanced skills in artificial intelligence (AI), he shares his insights on how his experiences in the USA have shaped his work in Sweden. He also highlights the challenges and opportunities Sweden faces within the life science industry and provides his perspective on how we can balance innovation with patient safety and privacy.</em></p><h4><strong>Can you tell us about your background in Sweden and the USA?</strong></h4><p>- I was born in Stockholm and moved to Helsingborg during my upbringing. I attended medical school at the Karolinska Institute, completed my doctoral studies, and specialized as a cardiac intensive care physician there as well. I spent two years as a postdoc at UCLA (University of California, Los Angeles). After my time in Los Angeles, I was recruited to Sk&#229;ne. I co-founded the Stem Cell Center in Lund and became a professor at Lund University. Simultaneously, I was the head of the cardiac intensive care unit at Sk&#229;ne University Hospital (SUS) in Lund. After working as the Medical Director at the Cardiac Intensive Care Unit in Lund for five years, I was recruited to Michigan State University and Spectrum Health, where I worked as a senior physician and cardiologist in the thoracic intensive care unit and served as Research Director for nearly 10 years. During this time, I also became a professor at the Van Andel Institute in Michigan, as well as at Stanford University and the University of Texas. During my time in the USA, I earned a degree in epidemiology from Harvard University, specializing in secondary analyses. Subsequently, I completed a three-month program in Artificial Intelligence at the Massachusetts Institute of Technology. With these qualifications, I worked on data lakes related to intensive care.</p><div class="pullquote"><p>&#8220;You don't need to understand the details of AI to benefit from it, but you must understand the results.&#8221; - Stefan Jovinge</p></div><h4><strong>What are the biggest differences in working with health data in Sweden compared to the USA?</strong></h4><p>- In the USA, there is a strict yet consistent approach to data lakes. Individual data requires permission to be used. What has allowed AI to advance so far is the process of handling data differently when it is pseudonymized. This enables the creation of datasets where identifiable data is removed, and through keys, an so-called honest broker, who is not part of the projects using the data, ensures that the dataset can continuously be analyzed and updated. Creating these pseudonymized datasets requires permission granted after review and if the data meets the legal requirements for pseudonymization with sufficient security. This is referred to as non-human subject data. These datasets do not require ethical approvals or consent from the patients from whom the data originates. For example, a patient having a blood pressure of 125/75 mmHg requires the patient's permission to be used, but not 125/75 without a connection to the individual. This allows the use of complete datasets, which is currently not possible in Sweden.</p><h4><strong>How do your experiences from the USA influence your work in Sweden?</strong></h4><p>- I learned everything I know about large health datasets and advanced analyses of these in the USA. The Americans have solved the legal question, and this inspires us in Europe, where we now have the opportunity to locally adapt the EHDS and AI Act to create the same conditions that North America has, within the EU and Sweden.</p><h4><strong>You have extensive experience with AI and data lakes from your time in the USA. How do you view the development of these technologies within Swedish healthcare?</strong></h4><p>- Sweden has begun to understand the seriousness, and the EU is driving Sweden forward. Risks are starting to be identified. The development has finally led to the understanding that all data management involves risks, but we cannot stop all handling of large datasets because there is a risk&#8212;everything entails risks. For example, we do not shut down all traffic because 350 people die annually.</p><h4><strong>How can we balance the need for large datasets for AI development with the protection of patient privacy?</strong></h4><p>- Here we can learn from North America: they talk about alternative risk. When running, for example, a pilot project for the use of AI, one can calculate the benefits it brings if used on a large scale. For instance, how many breast cancer patients' lives are saved by the AI algorithm. By weighing this alternative risk against the privacy risk, a balance can be found. Our biggest privacy risks lie with our current electronic medical record systems. These data are often unencrypted and without pseudonymization. We happily use these systems today.</p><h4><strong>You are a member of an advisory group for life science to the Swedish government. What issues do you consider most urgent to strengthen Sweden's position in this field?</strong></h4><p>- We discuss many issues, but the use of health data is something I have driven. Our quality registers have given us an international reputation, and we still want to be an important player here. If our research and industry are to remain at the forefront, we must be able to make complete datasets available in pseudonymized form for analyses&#8212;otherwise, we will lose our position here. Life science is our third-largest export product, and it is extremely important from a national strategic perspective.</p><h4><strong>How can legislation be adapted to support innovation while protecting patient safety and privacy?</strong></h4><p>- We need to make complete pseudonymized datasets available that do not require patient consent. AI methodologies have enabled the discovery of more complex relationships than traditional statistics, but this also means that AI can more easily be misled by biased data. So, if you ask whether you can use patients' data, those who approve it will constitute biased data that can mislead AI.</p><h4><strong>How do you view Sweden's role in the global life science industry, and what steps are needed to promote innovation within the sector?</strong></h4><p>- Our actors (universities, companies, and academic hospitals) must be able to act as freely as possible, and it is crucial that they become attractive to international partners. It is through international collaborations that we can make a difference and be part of the international front line. Attempts to create "purely Swedish" solutions have been a Swedish theme that risks isolating us and causing us to fall behind.</p><h4><strong>If you could change one thing within the medical research field in Sweden, what would it be and why?</strong></h4><p>- Sweden must invest significantly more and not pour money into groups that can distribute it freely. All investments need to be reviewed by international expert panels to ensure that the funds have a chance to become beneficial.</p><h4><strong>Is there a particular project or research area you are especially enthusiastic about right now?</strong></h4><p>- Yes, three things.</p><ol><li><p>I believe that the AI initiative with the help of AI Sweden enables very good collaborations and projects.</p></li><li><p>Projects around &#8220;AI on AI,&#8221; where AI is used to explain predictions to the user about what drives the prediction, are something we are interested in and I am fascinated by. You don't need to understand the details of AI to benefit from it, but you must understand the results.</p></li><li><p>The work on common standards for AI is beginning to take shape, and I believe it will be crucial for AI to be used as safely as possible. We hope for no setbacks due to misuse.</p></li></ol><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.informerad.se/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Informerad.se! Subscribe for free to receive new posts and support my work.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><p><em>Note! The content on this blog reflects my personal opinions and does not represent my employer. As the publisher, I am not responsible for the comments section. Each commenter is responsible for their own posts.</em></p>]]></content:encoded></item><item><title><![CDATA[AI Pioneer Göran Lindsjö: From Linköping to the White House – A Journey Through the World of AI]]></title><description><![CDATA[With over 40 years of experience in the field, the AI nestor G&#246;ran Lindsj&#246; shares his insights on how AI has evolved and impacted both businesses and societies globally.]]></description><link>https://www.informerad.se/p/ai-pioneer-goran-lindsjo-from-linkoping</link><guid isPermaLink="false">https://www.informerad.se/p/ai-pioneer-goran-lindsjo-from-linkoping</guid><dc:creator><![CDATA[Johan Åhlén]]></dc:creator><pubDate>Sun, 15 Sep 2024 12:57:54 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!DXbr!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2169b652-b789-428c-8e30-5c5f176ab044_3504x2336.jpeg" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!DXbr!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2169b652-b789-428c-8e30-5c5f176ab044_3504x2336.jpeg" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!DXbr!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2169b652-b789-428c-8e30-5c5f176ab044_3504x2336.jpeg 424w, https://substackcdn.com/image/fetch/$s_!DXbr!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2169b652-b789-428c-8e30-5c5f176ab044_3504x2336.jpeg 848w, https://substackcdn.com/image/fetch/$s_!DXbr!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2169b652-b789-428c-8e30-5c5f176ab044_3504x2336.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!DXbr!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2169b652-b789-428c-8e30-5c5f176ab044_3504x2336.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!DXbr!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2169b652-b789-428c-8e30-5c5f176ab044_3504x2336.jpeg" width="1456" height="971" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/2169b652-b789-428c-8e30-5c5f176ab044_3504x2336.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:971,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:1581039,&quot;alt&quot;:&quot;G&#246;ran Lindsj&#246;&quot;,&quot;title&quot;:null,&quot;type&quot;:&quot;image/jpeg&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="G&#246;ran Lindsj&#246;" title="G&#246;ran Lindsj&#246;" srcset="https://substackcdn.com/image/fetch/$s_!DXbr!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2169b652-b789-428c-8e30-5c5f176ab044_3504x2336.jpeg 424w, https://substackcdn.com/image/fetch/$s_!DXbr!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2169b652-b789-428c-8e30-5c5f176ab044_3504x2336.jpeg 848w, https://substackcdn.com/image/fetch/$s_!DXbr!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2169b652-b789-428c-8e30-5c5f176ab044_3504x2336.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!DXbr!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2169b652-b789-428c-8e30-5c5f176ab044_3504x2336.jpeg 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container restack-image"><svg role="img" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg"><g><title></title><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button tabindex="0" type="button" class="pencraft pc-reset pencraft icon-container view-image"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a><figcaption class="image-caption">G&#246;ran Lindsj&#246;</figcaption></figure></div><p><em>In a time when artificial intelligence (AI) is rapidly transforming both businesses and societies, few have as long and broad experience in the field as G&#246;ran Lindsj&#246;. With a career that began in the 1980s, when AI usage in Sweden was still in its infancy, G&#246;ran Lindsj&#246; has helped shape its development both nationally and internationally. From working on the commercialization of AI tools in Sweden to nowadays advising some of the world&#8217;s most influential organizations and decision-makers, G&#246;ran Lindsj&#246; has witnessed how AI has changed everything from welfare to business models. In this interview, he shares his experiences and insights about AI&#8217;s role in our future.</em></p><h4><strong>You have a very long and solid background within AI. Can you tell us how it all started?</strong></h4><p>When I was a teenager, I read a bit about AI in the school library in Mj&#246;lby, but also the so-called bicycle issue of <em>Scientific American</em> in 1973. There was an article describing how humans were inefficient compared to several other animals in terms of energy use for movement. But humans had the ability to create tools, such as a bicycle, to become the most energy-efficient creature. It was fascinating to later see Steve Jobs describe how he was influenced by the same article. Combined with my interest in the logical aspects of AI, this led me to want to contribute, in some small way, to the reasoning aspects of creating new tools. At 17, I described a mathematical game that was published in <em>Scientific American</em> the following year and was actually put to some use.</p><p>A few years later, in the mid-80s, Swedish AI research was concentrated in Link&#246;ping. Several of the researchers started the company Epitec to commercialize their results. I applied to work there because I was intrigued by the way logic was handled, and above all, fascinated by the applications we worked on. We had an incredibly fun time for several years, and it shaped my view early on about how fundamental AI could improve business and welfare.</p><h4><strong>Can you tell us about your international AI experiences?</strong></h4><p>I started working internationally early on. Initially, it was with the export of our Swedish AI tools, and later with American AI solutions in the European market. In countless meetings in North America and Europe, I was infected by the American enthusiasm for AI&#8217;s possibilities. The Nordic countries were often ahead in applications compared to the rest of Europe. Much later, in the spring of 2016, we moved to Washington DC for various reasons. By then, the AI revolution had begun in North America, and it was incredibly exciting to be operational on the ground as AI changed the conditions in several industries. I had the privilege of working with the major AI companies, leading analysts, political advisors, and both the private and public sectors.</p><div class="pullquote"><p>&#8220;With AI, one can work broader and more on human terms. This means that many AI applications can be based on interaction with, for example, customers or citizens rather than being based on well-defined internal processes.&#8221; &#8211; G&#246;ran Lindsj&#246;</p></div><h4><strong>I know you've been to the White House in the US to discuss AI. Can you tell us about that?</strong></h4><p>Three of us from AI Sweden had the privilege of meeting with the Biden administration&#8217;s central AI office at the White House, as well as the president&#8217;s national security advisor, among others. We also had interesting discussions with, for example, the US Department of State. Everyone was very open to us, and perhaps the main takeaway was how necessary it is for the US, together with its allies, to lead the development of AI.</p><h4><strong>You&#8217;ve worked in the Swedish Government Offices. What lessons did you take away from there?</strong></h4><p>I worked, among other things, on the digitalization of major agencies. The Government Offices are a well-oiled machine for cooperation with the EU and for implementing political decisions. It is more operational than strategic or advisory to the government. In recent years, I&#8217;ve had the opportunity to work more closely with some Western counterparts to the Government Offices and often feel they have a higher strategic level.</p><h4><strong>Which regions and countries have advanced the most in AI? How do Sweden and Europe compare?</strong></h4><p>According to various rankings, which state that the US leads followed by China, I believe this holds true in all essential areas. The UK, Canada, Singapore, South Korea, and Israel are also far ahead, though it varies across different parameters. Summarized by larger regions, it&#8217;s the northern parts of the Americas and Southeast Asia that lead. The EU lost more than half of its AI capacity with Brexit and now faces an uphill battle.</p><h4><strong>You work as an AI advisor and educator to politicians and leaders of large organizations and companies. What do you see as the most important knowledge for politicians and leaders regarding AI?</strong></h4><p>Politicians primarily need to understand the power of AI as a tool to implement a political agenda. This could be in welfare, energy, security, climate, and much more. When politicians believe AI is something technical, the administrative leadership in agencies, regions, and municipalities often limits AI to efficiency improvements and continuous refinements &#8211; focusing inwardly. An exception is, of course, medical research. In the private sector, it&#8217;s more varied. Large export companies have generally understood for several years the power of AI in developing business operations. At the same time, there are boards that have yet to grasp AI&#8217;s disruptive power in their industry. Similarly, in the public sector, this often results in the operational management limiting AI use to efficiency improvements. As a result, they risk competitors incorporating AI into services or products, or even worse, new players using AI for new business models. For politicians, it&#8217;s about AI as a political tool, and for companies, it&#8217;s about AI as a business tool.</p><h4><strong>What would you say are the biggest differences between AI and regular digitalization?</strong></h4><p>Digitalization deals with well-defined problems in a structured world. You know in advance how the processes look; input, algorithm, and output. This applies to, for example, payroll software in an HR department. When reality isn&#8217;t as orderly&#8212;which it rarely is&#8212;especially when dealing with people, digitalization typically requires humans to adapt to the machine: "Press 4 for a mortgage." With AI, you can work more broadly and on human terms. This means that many AI applications can be based on interaction with, for example, customers or citizens rather than well-defined internal processes. One of the most common mistakes is therefore treating AI as a subset of digitalization, which is unfortunately particularly common in Sweden.</p><h4><strong>What advice would you give to today&#8217;s youth considering the development of AI?</strong></h4><p>I think of Kevin Kelly, who said a few years ago that the next 10,000 companies will all be X+AI. That is, they will focus on various X&#8217;s, such as a product in HR or a service in elderly care, and then combine it with AI to create something unique. He has been mostly correct. Perhaps young people can think similarly: X+AI, where X is my interest or what I want to deepen my knowledge in. It could be economics, law, sports psychology, or anything else. But combine it with AI. Sweden has a significant shortage of such individuals, possibly with the exception of areas where X is technology or natural sciences. Unfortunately, in Sweden, AI has been positioned as something technical in the education system, meaning there are limited programs in non-technical fields offering AI elements.</p><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://www.informerad.se/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Thanks for reading Informerad.se! Subscribe for free to receive new posts and support my work.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div><div><hr></div><p>Note! The content on this blog reflects my personal opinions and does not represent my employer. As the publisher, I am not responsible for the comments section. Each commenter is responsible for their own posts.</p>]]></content:encoded></item></channel></rss>