MethodKit

Everyone's racing to connect AI to their company's knowledge. Almost no one stops to ask where that knowledge actually is. Most of it is dark. It was never written down. It lives in someone's head, or their hands, and it walks out the door the day they leave. Andriy Z. and I mapped the whole journey: from knowledge nobody can even put into words, to knowledge an AI acts on before you ask. Seven levels, three colors. Dark, gray, green. The catch: most teams are attempting level 7 (let the AI act) on top of level 1 context (it's all still in people's heads). You can't automate what you haven't written down yet. The 7 context levels 👇🏼 https://lnkd.in/dSMwJ2Xd

Understand what you do, so AI can help you do it better. We've built an AI workshop that starts with you, not with the technology. Working with AI takes two things. The tool: sitting down at the computer and learning it. And the context: the setting you work in. Almost all training covers the first. This workshop focuses on the second. That's why John Airaksinen and I are running a half-day in his salon in Hornstull. We start from your reality, make it visible with MethodKit + unfolding Dark Context, and build AI prototypes based on it. John and MethodKit overlap a lot in how we think. We both see that context, and an understanding of what you actually do, is the key to making the technology work well. We make how you work visible with cards and a clear visual overview. John works with the salon at the center, where experiences and spaces come first. In the workshop, John & Julius Salong (in Stockholm) becomes part of the method. The room and the conversation become a prompt in themselves, drawing out the things you'd otherwise never put into words. More info at (in Swedish): https://lnkd.in/dJ7GKV3V

We're working on a new project called Dark Context. Explore more at https://darkcontext.ai/ Ola Möller, Andriy Z., Dark Context

SOMETHING NEW WE'RE WORKING ON: For decades, many companies never bothered to write down what they know. AI just put a price on it. Every organization runs on things that live in no document. - How decisions really get made. - The workaround that's in no manual. - The judgment a new hire needs a year to absorb. Invisible, and invisible felt free. For years the answer to "we keep losing knowledge" was simple: write it down. Wikis, handbooks, the shared drive, the onboarding doc. Most of it goes stale, half of it is never found, and the hardest-won knowledge never makes it in at all. Then the stakes changed. A general AI model (LLM) knows almost everything in general, and almost nothing about you. Feed it your context and it gets genuinely useful, fast. Feed it nothing and it stays confident, generic, and slightly wrong. The same gap that used to leak away quietly when someone quit now decides whether the most powerful tool of the decade works well for you, at all. That split is forming right now. A few organizations are turning what they know into something AI can use, and pulling ahead. Most haven't started, and the distance compounds. The prize was never a model that knows everything. It's an organization with an accessible memory. Andriy Z. and I will continue to add thoughts to https://darkcontext.ai/

🎙️MethodKit has joined the founding collection. Not every organisation in this collection has thousands of employees. Some are small. Some are built around a single idea and a network of collaborators who believed in what it could do. MethodKit is one of those. Founded in Stockholm by Ola Möller, it has spent over a decade making card decks that help teams think together. More than 200 collaborators – facilitators, translators, subject-matter experts – have shaped the 60+ decks that now exist. One of them has been translated into 43 languages. #MoOC #MuseumOfOrgCulture #Curatology #FoundingCollection

The AI conversation usually starts with the AI tools. We start it with the people. MethodKit for Journalism & AI gives a newsroom a shared language to map its own workflows, ethics, and risks before reaching for any tool. Built with DW Akademie and team of 15 journalists, AI experts and educators across 12 countries: Kenya, Namibia, the Philippines, Bolivia, Zimbabwe, Spain, South Africa, Germany, Ukraine, Qatar, Eswatini, and Sweden.