Marcus and Chollet were right: code execution as neurosymbolic AI is a top breakthrough

joshua_saxe · x · 2026-09-04

The author argues Gary Marcus and François Chollet have been under-credited for presciently positioning neurosymbolic reasoning and program synthesis as central to AI progress.

His core claim: giving LLMs the ability to execute code lets them "jailbreak the liminal space between reason and reflex" and manipulate the "platonic realm" directly — using programs as precise formal handles on abstract concepts. He ranks this among the top 3 AI breakthroughs of the past 4 years, drawing a parallel to humanity's adoption of computers since the 1940s as a similarly transformative leap.

Related event: Debate Erupts Over Whether Gary Marcus's Neurosymbolic AI Stance Was Prescient(5 posts)→

Original post →

More from AGI Musings

AGI Musings channel →