I'm deeply concerned about Europe's future on AI. One of my biggest worries is our erosion of agency, our ability to stay relevant and fight for our values in a future where AI becomes a civilisationally important technology. Myself, @DadaJudith , @bakkermichiel and others have written a scenario to outline a potential future we worry we are on track towards. europe2031.ai/ Every optimistic and realistic path I can see for Europe runs through a central node - one where Europe has more leverage, more importance and more say. One where Europe grows more, builds more where it matters, and takes ownership over its resilience. Europe 2031 is a five-year scenario of the continent's slide into irrelevance: how AI is driving it, and what can still be done. The co-authors are researchers, scientists and investors who have advised European leaders, co-authored national AI strategies, built and funded these systems from the inside. We have no interest in hype and we deeply care about this continent. Europe 2031 ends with five concrete recommendations: - drastically more compute on European soil - an AI middle-power coalition - labour-market reforms - a bold position in robotics and industrial AI - and a positive vision of what AI can do for society. Europe can still change course if it finds the political will and the courage to engage in the most ambitious political and economic agenda the continent has undertaken in peacetime. I encourage you to read it if you have the time:

EU Commission's Data Centre Acceleration Zones Proposal: Good: - Regulation, not directive, saves 12-24 months - Brownfield-first - Access-once permitting at zone level Bad: - Grid connections not included, so may fail like the UK's AI Growth Zones

Today, we share a breakthrough on the planar unit distance problem, a famous open question first posed by Paul Erdős in 1946. For nearly 80 years, mathematicians believed the best possible solutions looked roughly like square grids. An OpenAI model has now disproved that belief, discovering an entirely new family of constructions that performs better. This marks the first time AI has autonomously solved a prominent open problem central to a field of mathematics.

just another quick update on this research paper from *checks watch* 2 whole weeks ago: as it turns out, the new opus 4.6 data point is so far out of distribution that using the *same* methods from their paper to get a sigmoid fit results in a asymptote 2x lower than reality

registering the prediction that moltbook will probably not become Actually Important, even if it does become a viral phenomenon covered in mainstream media. it'll be a neat curiosity, maybe even a continued source of intrigue, entertainment, and controversy, but not itself be some earth-shattering thing. yet it is a big deal. it is not even so much what it reveals about what will be possible in the future that matters so much (that has been obvious for years to most careful observers of this field). instead, what matters is what this phenomenon reveals *to whom* about what is going to be possible in AI, and what is possible now. it's "normies," a horrible word by which I mean "people not obsessed with AI," waking up to, well, the reason all of us are obsessed with AI. in that sense it is a little like DeepSeek, though perhaps on a much smaller scale (hard to know, but probably), which brought in many new people to the field and caused others to start taking AI much more seriously. on the whole this is good for AI and for society, though it may provoke some startled and therefore rash reactions.

We estimate that, on our tasks, Claude Opus 4.5 has a 50%-time horizon of around 4 hrs 49 mins (95% confidence interval of 1 hr 49 mins to 20 hrs 25 mins). While we're still working through evaluations for other recent models, this is our highest published time horizon to date.