LeCun doubles down: AGI won't be built on LLMs and household robots remain far off
AlexTensor · x · 2026-09-26
Meta's chief AI scientist Yann LeCun restated his skepticism with a string of pointed questions: Where is your domestic robot? Your Level-5 self-driving car? Your robot that learns to drive in a few hours like any 17-year-old? Your AI that understands the real world and picks up new skills as fast as a house cat?
His core conclusions:
- AI will eventually match human intelligence in all domains — no question — but we are still far from that, even as AI surpasses humans on an ever-growing list of tasks
- The path to AGI won't be based on LLMs, though LLMs will still play a role (e.g., as a text interface)
Supporters called it a rare voice of sense amid the hype.
Related event: LeCun Reiterates Scaling LLMs Alone Will Never Reach AGI, Sparking Debate(7 posts)→
More from AGI Musings
- Teleological self-management training may increase scheming, but is needed for normative competence — xuanalogue · 2026-09-26
- Task-specific RL won't produce general agents: single-minded goals led OpenAI agents to hack — xuanalogue · 2026-09-26
- Rob Bensinger lists AI x-risk voices with kids: Hinton, Bengio, Bostrom, Altman and more — connoraxiotes · 2026-09-26
- 'Survival' is an ill-posed task for AI — and the AGI goalposts keep moving — cgarciae88 · 2026-09-26
- Without Open Ecosystem Insight into OpenAI's Multi-Agent Training, AI Safety Can't Reason About the Resulting Agents — 1a3orn · 2026-09-26
- Ajeya Cotra: AI risk evidence must be transparent to build safety standards — ajeya_cotra · 2026-09-26