Tal Linzen: LLM 'Reasoning' Means Following Algorithms, Which Fails as Complexity Rises

tallinzen · x · 2026-08-05

Weighing in on the debate over whether large language models can truly reason, AI researcher Tal Linzen argues that the term has essentially devolved into meaning "following an algorithm exactly."

Under this definition, transformer-based models cannot genuinely reason via chain-of-thought. Linzen points out that empirical evidence repeatedly shows models making increasing mistakes as problem complexity rises, even for simple algorithms.

Related event: Researchers Debate LLM Reasoning Limits as RELIC Framework Exposes Failures(6 posts)→

Original post →

More from AGI Musings

AGI Musings channel →