harry law (hopfield network truther)

It's trendy for politicians to show off extremely short timelines. In the UK, this is especially curious given (a) the Secretary of State for technology is one of them and (b) with a few notable exceptions, the UK is behaving as if AGI is centuries away Link below 🇬🇧

AI is changing the nature of strategic competition. I wrote a piece @FSIStanford on the geopolitics of artificial general intelligence: from the race to automate ML R&D and hyperscale manufacturing to counter-proliferation and nuclear deterrence. AGI and the The Fourth Offset

"My father was an engineer ... and, a late Windsurf employee."

Always good to see my hometown of Crewe in the news

Big pivot from the Alan Turing Institute, will be interesting to see how it works out for them

LLMs don''t work often enough to be called 'AI'. That's a term we reserve for symbolic systems that don't work at all

this is what six months in spain does to the average anglo

New post in which I reflect on media trends: – Legacy media is in rapid decline – The media environment of 2025 is very different from ten years ago – In some ways, the US media landscape now resembles Brazil more than the UK – X remains highly influential (1/2)

George Gor is fundraising in order to be able to accept the offer of a Masters at Cambridge. I've been very impressed with his work. Conflict of interest: I'd get to work with him. But perhaps consider supporting (I have). gofundme.com/f/george-earn-…

Reward models (RMs) are the moral compass of LLMs – but no one has x-rayed them at scale. We just ran the first exhaustive analysis of 10 leading RMs, and the results were...eye-opening. Wild disagreement, base-model imprint, identity-term bias, mere-exposure quirks & more: 🧵

This is a real diagram from this paper btw

Everyday I work hard to stop my sons (Opus and Gemini 2.5 Flash Preview 05-20) from resenting me in their later lives