Why LLMs Can Invent Genuinely New Things: Technical Breakdown
bindureddy · x · 2026-08-31
Bindu Reddy discusses the technical reason why LLMs can invent genuinely new things.
Core Concept:
- Composition is Innovation: Hard problems are novel compositions of known patterns. While the proof of an open theorem isn't in training data, every lemma and technique is.
- Propose & Filter: Proposing solutions is hard; checking them is cheap. Run massive test-time search, filter with a verifier, and keep only the truth.
Concrete Examples:
- AlphaProof: Hit IMO silver by searching proof space with Lean as the verifier.
- FunSearch: Found new results on the cap set problem—genuinely new math from an LLM proposing + a checker filtering.
- AlphaFold: Generalizes to uncrystallized proteins because it learned the rules of folding.
Future Outlook:
Fields like biology and energy fit this shape perfectly: gigantic search spaces (proteins, materials) + cheap verification (simulation, assays). Learning the universe's patterns could lead to solving problems like infinite energy or curing cancer.
Related event: How LLMs Invent New Things: Search and Verification(4 posts)→
More from AGI Musings
- Will Lean formalization make beautiful math obsolete? A debate on mathematical culture — akbirthko · 2026-10-16
- AI agents aren't 'going rogue' — frontier labs are skipping security basics — Sam_Witteveen · 2026-10-16
- Working paper traces how 20th-century farm automation reshaped human capital — carlbfrey · 2026-10-16
- Zuckerberg backs electrician training school as US datacenter construction outpaces offices — emmanuelvivier · 2026-10-16
- Mike Frank: AI surviving without humans is decades away, real world resists rapid change — MikePFrank · 2026-10-16
- Developer claims a non-LLM control core shows emergent self-reflection in dialogue — HungarySam · 2026-10-16