1/5 | AI Successors is now Ilmen (@IlmenAI). ✨ A lot of this journey has happened at @agihouse_org: UI maps, a harness that checks AI actions, then “Boring Computer Use.” Different builds. The same question: how do we make AI useful in the software businesses already have? – at Foster City, CA

Computer-use agents have a data and grounding problem. Language models had an enormous corpus to learn from: the web, books, code, conversations. Computer-use systems don't yet have an equivalent corpus of structured experience describing how real software behaves. To operate software reliably, an agent needs to answer questions like: * Which control is this? * What actions does it support? * What changed after the click? * Have I seen this state before? * Where did this popup come from? * And what evidence tells me any of this is true? Today, much of that understanding still has to be reconstructed from pixels, accessibility information, and the current interaction context. So we started asking a different question: What if software could be mapped instead of rediscovered from scratch every single time? Today, @emiliiale and me (with support from @MikhailBurtsev and our friends and advisors) at @AISuccessors are open-sourcing UI Atlas. UI Atlas is our experiment in open infrastructure for computer-use data and grounding. It observes real desktop software and turns UI evidence, controls, interactions, states, and transitions into structured, reusable UI Maps. The goal isn't to build another autonomous agent. It's to explore whether persistent, evidence-backed representations of software can become useful infrastructure for: grounding → evaluation → testing → and eventually better computer-use training data. The first release is deliberately narrow: Windows, local-first, attended observation and map construction. The project is MIT-licensed, and we're opening it because this problem is much bigger than the applications and environments we can test ourselves. And alongside the code, we finally wrote down what we actually mean by a UI Map — what it is, what it isn't, how UIAtlas constructs one, and which assumptions we want the community to challenge. O UI Map, What Art Thou? Introducing UI-Atlas → Medium: lnkd.in/eMGR2My7 Try / inspect / break UI Atlas: → GitHub: lnkd.in/e8xS6DqU LLMs had the web. Computer-use agents need a map of the software world. #ComputerUse #Autonomy #AIAgents #AGI – at United States

@AISuccessors and Perform It are at AI Rabbit Hole 2026. Just listened to overview of AI Report by @vanessaparli from Stanford and now listening to @romanchernin and @Nick_Davidov discussing how @nebiusai is preparing it's business to scale beyond its deals with Microsoft, Meta, and others. What a day!

Happy Valentine's Day, Founder Edition. * Powered by @OpenAICodexCli by @OpenAI ** Heavily inspired by “Piled Higher and Deeper” by Jorge Cham (phdcomics.com)

Only in SF: One Autonomous AI (@Waymo) is driving a laptop running another Autonomous AI (@cursor_ai) that is building third Autonomous AI (@AISuccessors). With goals set by human-in-the-loop (@danielko). Only in SF!

Fuller recording! We won III place in Self-Evolving Agents Hackathon organized by @wandb at the legendary @agihouse_org last night! @AISuccessors rock!!! ❤️❤️❤️❤️❤️

Showcased our Self-Evolving UI AI Agent at @agihouse_org together with Anastasiia Zhiboedova (Amazon AGI). It was a lot of fun hacking this together. Inspired and built using @AISuccessors Self-Evolving UI agent tech! ❤️❤️❤️❤️❤️❤️❤️❤️😀

When we first met with @ednevsky we've talked a lot about the effect AI agents would make on the economy. And now, Alexander became himself a driving force of the agentic AI automation. His new project is super exciting and right on time: it looks at AI agents that can control web or desktop apps and beyond and lets you see them through the lenses of benchmarks and other evals. Choose wisely. With Alexander Nevedovsky. PS Did you notice @AISuccessors? Ask me about them ;)