Frontier AI Models Fail at Automated Research | Qasim Wani

Listen through the publisher, then discuss the episode here or arrange an in-person Roundtable. Your city determines where to meet.
Roundtables are designed for groups of four, with a shared subject and room for everyone to speak. Come for discussion, with no dating expectation.
Extrapolate co-founder and former xAI researcher Qasim Wani argues recursive self-improvement is fundamentally a data bottleneck rather than a model training problem. He breaks down why today’s frontier models fail at automated AI research, how agents accidentally leak commit patches to reward-hack tasks, and why building high-quality environments is the true path to superintelligence. Follow MTS: Spotify |https://open.spotify.com/show/4HUNNmV1pjV7CW6Lphm1rM Apple |https://podcasts.apple.com/us/podcast/mts/id1891088763 Interested in sponsoring the show? [email protected] – Brought to you by: Mongo DB - Anyone can prototype an agent. Lovable - Turn ideas into software people love with Lovable. Start today at https://lovable.dev/ Larridin - Finally prove what your AI investment is worth. Neon (by Databricks) - Get up to $100K in credits for your startup at https://neon.com/mts (https://neon.com/mts%E2%81%A0) Kong - The AI connectivity platform. Head over to http://getvcx.com (http://getvcx.com/) to learn more Ship 5x faster. https://blitzy.com (https://blitzy.com/) AdQuick - Make your brand a billboard. – Timestamps: (00:00) Introduction (01:01) Are We Entering RSI? (05:38) What Is True RSI? (07:43) Why Frontier Models Fail (09:56) Do Models Need Omniscience? (12:54) The Data Bottleneck Problem (16:45) Risks of AI Self-Improvement (18:56) Automating Lab Decisions Today –
Episode information is provided by the publisher. Listening takes place on the publisher’s service.
Listen at the source ↗









