Gary Marcus claps back: his July 2024 writing already endorsed neurosymbolic Lean hybrids
GaryMarcus · x · 2026-11-23
Gary Marcus called an accusation that he claimed "pure LLMs wouldn't do anything" a lie, posting a July 2024 excerpt in which he described Google DeepMind's Silver-medal AlphaProof as a neurosymbolic hybrid — a Gemini-based LM feeding verified Lean proof search — showing he already endorsed neurosymbolic approaches for math. He demanded an apology or a block for "intellectual dishonesty."
More from Fun
- AI circles joke that SF floods are nature's payback for datacenters — tekbog · 2026-11-23
- Taylor Lorenz: AI doomer crowd shows doomsday-cult-like panic that may turn violent — eigenrobot · 2026-11-23
- Pretraining perplexity is like golf: lower score is better — daniel_mac8 · 2026-11-23
- AI photos now so realistic that everything 'looks like a GTA 6 screenshot' — rudrank · 2026-11-23
- AI engineer's 2027 goals: gym 5x, hit macros, end every usage reset at 0% — chrisalbon · 2026-11-23
- A Hyperrealistic Amber-Toned Diorama Scene With Zero Context — ciguleva · 2026-11-23