Debating RSI, the US-China Gap, and Jaggedness with JS Denain of Epoch AI
Epoch AI's JS Denain and Nathan Lambert debate RSI evidence, the US-China gap, and existential risk.
Nathan Lambert interviewed Epoch AI insights lead Jean-Stanislas Denain about recursive self-improvement, the US-China model gap, and capability jaggedness. Denain said public OpenAI and Anthropic evidence, including roughly 2x monthly growth in researcher Codex spending, does not strongly show an imminent software intelligence explosion. He remains highly uncertain but considers roughly 10% existential risk over a decade reasonable. The discussion also covers robotics, distillation, Chinese lab hiring, open versus closed models, and frontier post-training.
- Denain says published lab metrics do not prove imminent self-sustaining AI acceleration.
- OpenAI reportedly saw roughly 2x monthly growth in researcher Codex spending.
- He treats about 10% existential risk over a decade as plausible but uncertain.
- Topics include Chinese model lag, distillation, and open versus closed safety.
- Countries
- United StatesChina
From the original article
16 min read at interconnects.aiThis episode is with Jean-Stanislas “JS” Denain of Epoch AI, who leads their Insights Team and is one of the people I find myself debating the state and trajectory of AI with more and more. We’ve had follow-on discussions of many of my favorite recent posts online and/or in private, so I wanted to dig into the nuance in a public episode.
By Nathan Lambert · interconnects.ai
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.