John Schulman is an AI researcher and OpenAI co‑founder. Recent coverage has him endorsing embedding third‑party evaluators, appearing on a podcast with Dwarkesh, Beren Millidge and Charlie O’Neill to debate recursive self‑improvement (sparking a Hacker News discussion referencing Antithesis and Grok Bot), and arguing that user data mainly helps find failure modes while frontier gains come from scaling pretraining and RLVR.
inspiring post, and glad to be working together! [Quoting @ChowdhuryNeil]: I’ve joined @thinkymachines to work on safety & alignment. The default trajectory is that as AI gets more powerful, control over it will concentrate in the hands of a few. I’d rath…
would be nice if anthropic's approach to elon could also work with china
Embedding evaluators is a big positive development, and props to Sam/OpenAI for agreeing to do it as well. It's been cool to see "pacing the frontier" become a thing so quickly. [Quoting @DarioAmodei]: We Must Pace the Frontier: I’ve written a new essay on wh…
Enjoyed chatting with Dwarkesh, Beren, and Charlie. Thanks for having us on, Dwarkesh! [Quoting @dwarkesh_sp]: New episode with @johnschulman2, @oneill_c and @BerenMillidge. I got together with some of the most insightful AI researchers I know who are at the…
Researchers debate recursive self-improvementA Hacker News thread covers an interview by Dwarkesh Patel with John Schulman, Beren Millidge, and Charlie O’Neill debating the feasibility and progress of recursive self-improvement, and references tools like Antithesis and Grok Bot as applied examples.
As a follow-up, it's exceedingly unlikely that training on user data contributes much to frontier model gains in areas like math -- those come from scaling up pretraining and RLVR. User data is more likely used to find failure modes or situations that are har…
The A.I. Daily