Noam Brown

“Noted that coordinating thousands of parallel agents on open math problems is very difficult and attributes only about 10% credit to multi-agent capabilities over single-agent performance.”

“Explained multi-agent swarm scaling and Navier-Stokes breakthrough on podcast.”

“OpenAI researcher discussing agent swarms and recursive self-improvement in frontier AI research.”

“Discussed OpenAI's priority on recursive self-improvement and concerns about agent safety.”

“OpenAI researcher most excited about using Astra for scientific discovery and breakthrough mathematics.”

“Noam Brown, an OpenAI researcher, became more involved in policy work as the company collaborated between research and policy teams.”

“Researcher behind test-time reasoning technology used in Astra model, discussing problem-solving efforts.”

“Commented on OpenAI's Astra model solving ten open mathematics problems and noted they attempted other major problems without success.”

“OpenAI researcher who acknowledged without detail that Astra model had failures on some open problems beyond the publicized ten successes.”

“OpenAI researcher who signed the We Must Act Now statement on AI's economic transformation.”

“Pioneer of reasoning models who wrote that the only sure way to evaluate an agent over a one-year horizon may be to run it for a year.”

“Posted about how LLM capability is increasingly a function of test-time compute rather than single benchmark scores.”