Gary Marcus: I wrote 2000 posts on neurosymbolic AI over 25 years, but mocked for one omission
GaryMarcus · x · 2026-10-09
Gary Marcus complained that across 25 years he wrote roughly 2000 tweets and articles (including "Deep Learning is Hitting A Wall") arguing for supplementing deep learning/LLMs with neurosymbolic AI — yet "tech bros" are circulating the single instance where he forgot to mention it, ignoring the rest of his body of work.
The spat is the latest round in the long-running debate over whether LLMs are hitting a wall.
More from Fun
- Filming with an infrared camera to become a talking cartoon duck, made with Luma AI — mrjonfinger · 2026-10-09
- LLMs play Connect 4 via robot arms in MuJoCo; fable 5.1 wins at $17.40 in API calls — eigenron · 2026-10-09
- Multicellular 'alien life' spontaneously emerges from pure random noise — Promptmethus · 2026-10-09
- Why hard workers hate AI: fear of a 'Brahmin caste' of Idea Guys — menhguin · 2026-10-09
- Mathematicians 'outraged' as AI encroaches on their status, joked on X — miniapeur · 2026-10-09
- Banned From an AI Model for Abusive Behavior, Forced to Write Code by Hand Again — Dan_Jeffries1 · 2026-10-09