Gary Marcus: AGI unlikely this decade, foundational problems remain
GaryMarcus · x · 2026-08-25
In a reply about the timeline for AGI, Gary Marcus stated it is certainly not happening this decade. He pointed out that many foundational problems, such as continual learning, RSI, world models, and neurosymbolic breakthroughs, remain unsolved. He considers 2031 possible but unlikely, projecting AGI might arrive sometime in the 2030s.
Related event: Gary Marcus Says AGI Won't Arrive This Decade(2 posts)→
More from AGI Musings
- Human taste can be gamed into optimized slop; preference is not a verifiable reward — torchcompiled · 2026-08-25
- Grove Research Founders Discuss Hugging Face Incident and Multi-Agent Dynamics — Miles_Brundage · 2026-08-25
- AI Should Exceed Benchmarks, Exploring New Frontiers of Human Potential — manosaie · 2026-08-25
- Viewpoint: Being Anti-AI Is Being Anti-Cancer and Longevity Research — justalexoki · 2026-08-25
- Writing Feedback Bottleneck Is Over: AI Changes Education — paulnovosad · 2026-08-25
- AI is the only major growth engine; without it, global outlook would be dramatically worse — iamtrask · 2026-08-25