Gary Marcus Argues LLMs Can't Achieve Human-Level Math Without Symbolic Operations
AndrewLampinen · x · 2026-07-24
In a debate over LLMs' mathematical capabilities, Gary Marcus reiterates his long-standing argument that human-level cognition requires symbolic operations somewhere in the system. When pressed on whether models using external tools count as being 'good' at math, the discussion highlights the boundaries of neurosymbolic AI.
Related event: Gary Marcus Debates LLMs and Neurosymbolic AI Boundaries(4 posts)→
More from AGI Musings
- Open-network research can reveal how a place looks to machines, not people — schwentker · 2026-07-24
- Conference slide says Graeme Segal is the mathematician AI will replace last — AlexKontorovich · 2026-07-24
- Charliermarsh: It's Worth Learning Things Useful Today Even If Obsolete Tomorrow — charliermarsh · 2026-07-24
- A new AI-era bottleneck is the physical world, not idea generation — xiao_ted · 2026-07-24
- A model screenshot argues the real problem is turning intelligence into agency — MoonL88537 · 2026-07-24
- AI’s underrated third use case may be contract formalization and negotiation — curious_vii · 2026-07-24