New piece! 👽 Diffused AGI "advocate agents" could allow us to solve intractable social and political coordination problems. This same technology offers a foundation for better governance, allowing us to rebuild decaying institutions from the ground up. Link below!

It’s not about the children. Never has been.

To make progress on multi-agent safety, we should study the properties that predict agent behaviour in cooperation problems. Read our recent blogpost, by @tilli_cecilia, on what key agent properties we should focus on, and more on why this matters. Link below.

Great to see @WhiteHouse and @ENERGY initiative recognising AI’s incredible potential to accelerate progress in science - excited to collaborate on this!

Ha, ha, but no. Baumol 101: in a free market for labor, workers share the gains from the most productive sectors. Wages rise everywhere. Sectors that get productive more slowly—teaching, nursing, surgery—get pricier. Not because patients are worms. Because arithmetic.

This is an absurd lie which undermines substantive efforts to improve opportunities for low income students. Academic standards *have* catastrophically slipped. When you lie like this you destroy all your credibility on the topic.

From the makers of the popular AlphaGo documentary, The Thinking Game gives a much broader picture of the story of DeepMind and our mission to build AGI, drawing on interviews with myself and others going back many years. You can now freely watch it here:

I hate that AI has made me approach cute animal videos with an air of suspicion. I resent having to harden my heart to the possibility of an unlikely friendship between a very large dog and a very small one

Good explanation as to why I found the Yudkowsky/Soares book completely unpersuasive: https://t.co/EMrzI1ghrW

In "Coasean Bargaining at Scale" (Link below) Séb Krier (@sebkrier) argues that AGI should not be understood as a centralising, sovereign intelligence but as a technology capable of collapsing the transaction costs that historically justified large, hierarchical institutions.

BREAKING: 🇪🇺 EU to monitor private chats The EU reportedly plans to pass the controversial "Chat Control" legislation without discussion behind closed doors - MEP Martin Sonneborn reports on X

@scaling01 nope! my weak-ish impression is that these take minimum of ~25 hours staff time, can be more like ~60 hours in rougher cases. some example time costs: - API issues - unclear ex ante which scaffold model will be most performant on in held-out dev task set - automatically +

Who's working on an LLM-based tool to identify the 100 most poorly written or obsolete laws and regulations in each jurisdiction? It identifies the problem, provides evidence of why it needs to be changed and suggests new wording that a legislator can review, propose and enact.

Witkoff still thinks it's about a bit of territory. From his leaked call with Ushakov. 'Now, me to you, I know what it’s going to take to get a peace deal done: Donetsk and maybe a land swap somewhere. But I’m saying instead of talking like that, let’s talk more hopefully

I'm not sure if there's an AI bubble, but right now I think the "no bubble" side has stronger arguments. People are betting that AI revenues are going to go up a lot over the next couple of years and that seems plausible to me.

Our critics say our work will destroy the world, and many now point to "If Anyone Builds It, Everyone Dies" as the canonical case for AI doom. Yet we find the book's arguments extremely weak. The book might make for interesting fiction, but it never presents any evidence.

Ilya sounding very Hayekian here. Many specialized AIs:

AI in the Machinery of Government | @deanwball with @hamandcheese, @sebkrier , and @fiiiiiist

Part of the reason the current moment--global conflict, rise of populism, resentment, etc--has been so surprising to most in the policy space is that they were trained in the standard economic thinking of the time, which emphasized material comfort above all else. The emphasis

americans don’t appreciate enough that their hare-brained blogposts might be harmless at home, but often get picked up months later in the imperial periphery by wide-eyed locals. this stuff is worse for the world than all the plastic straws you guys throw in the ocean combined

Fukuyama was so prescient. In a society with strong rights and material comfort, but light on demanding shared purposes and some degree of sacrifice, thymotic energies go searching. Some quiet into bourgeois hedonism; other will seek “metaphorical wars” and eventually real ones.