OpenAI Staff Rebuttal: Can LLMs Create Einstein-Style New Knowledge?
inductionheads · x · 2026-08-24
Yong Zheng-Xin from OpenAI's RSI team published "Hot Take: LLMs Can Jump," countering papers that claim LLMs cannot make Einstein-style leaps. He argues in an interview that Feynman derived general relativity from existing theories without Einstein's original path.
Key points:
- Reasoning Types: Explains the specific reasoning type the paper claims LLMs lack.
- Acceleration Gap: Discusses why cyber capabilities accelerate while algorithmic progress seems stagnant.
- Generalist Advantage: Non-specialist models can find connections across fields that specialists miss.
- Continual Learning: Interconnected knowledge structures make continual learning easier.
More from AGI Musings
- DeepMind's Geoffrey Irving draws Mandelbrot in pure Lean via Claude, reflects on halting AI progress — DKokotajlo · 2026-10-16
- Mech Interp Veteran Rethinks Alignment, Misgeneralization, CoT Oversight — akbirthko · 2026-10-16
- GitCoin founder on Nick Land's technocapital: capitalism as a self-assembling machine intelligence — khademinori · 2026-10-16
- AI-driven job losses could turn $10T consumer debt into a $1T banking hole — churchkey · 2026-10-16
- Into the Unknown: an AI agent benchmark of 15 open research problems from US national lab scientists — Jeande_d · 2026-10-16
- Soon your AI will meet their AI — and humans without agents can't compete — danfaggella · 2026-10-16