Recursive co-founder Tim Shi on scientific taste, evals, and automating discovery
AnneliesGamble · x · 2026-11-18
Recursive AI co-founder Tim Shi discusses building recursively self-improving systems. Asked how AI might develop "scientific taste," he invokes deep learning history: many thought it a terrible idea, yet a small group persisted—proving the persistence worthwhile.
The core questions he tackles: when AI systems choose their own research directions, how do you recognize what's truly promising when near-term value is unclear? And how can AI learn from experts while still testing ideas those experts believe won't work?
His work focuses on the origins of scientific taste, eval design, and what it takes to automate scientific discovery.
More from AGI Musings
- AI Safety and Governance Work Often Pays Barely Above Subsistence, Insiders Say — JacquesThibs · 2026-11-18
- From Friendly AI to Anti-Handoverism: How Alignment Goals Drifted Toward Multipolarity — jessi_cata · 2026-11-18
- Three-minute compilation of AI CEOs admitting their tech could end humanity — zetalyrae · 2026-11-18
- Gary Marcus makes the case Anthropic isn't worth $1.4T, with Levie rebuttal to follow — GaryMarcus · 2026-11-18
- Interactive tree compares Paul Christiano's and Parzival's superintelligence timelines: 50% by late 2028 vs Sep 2027 — LeviTurk · 2026-11-18
- Fed warns AI agents could trigger bank runs in minutes — GaryMarcus · 2026-11-18