LeCun reaffirms autoregressive LLMs won't reach human-level AI, argues for search in continuous representation space
Scobleizer · x · 2026-09-21
Responding to Geoffrey Hinton's AI-danger campaign, Yann LeCun restates his claim that autoregressive LLMs alone won't lead to human-level AI. He argues current reasoning relies on non-autoregressive search done inefficiently in token space, while human-like reasoning should search a continuous representation space — a direction he says the industry is moving toward. He also notes current self-improvement methods only work where outputs can be scored without humans, such as math, code, and simulatable scenarios.
Related event: LeCun Reasserts Autoregressive LLMs Won't Reach Human-Level AI(6 posts)→
More from AGI Musings
- Agents' bottleneck isn't intelligence but goal commitment — and that's good for alignment — mimi10v3 · 2026-11-12
- Google Cloud survey: 78% say legacy enterprise vendors may not lead the AI era — rseroter · 2026-11-12
- Safety researchers blast OpenAI/Anthropic plan: automating alignment research with possibly misaligned AI — trevposts · 2026-11-12
- NYT essay: resisting Big Tech and AI takes collective action, not willpower — TuhinChakr · 2026-11-12
- Ethan Mollick: 'talk to your documents' RAG solutions already feel dated — emollick · 2026-11-12
- Anthropic's pitch hasn't changed in 5 years: 'AI is winner-take-all, we will build god' — menhguin · 2026-11-12