Gary Marcus: 2035 is already too generous for 'early 2030s' AGI claims
GaryMarcus · x · 2026-10-04
Gary Marcus doubles down on his critique of AGI timeline predictions, arguing that careful analysis is a lost cause against hype on X while praising @stevehou for trying. He adds that 2035 is overly generous to the original claim of AGI arriving in the "early 2030s" — the claim itself is even more aggressive.
More from AGI Musings
- AI Is Shifting the Cost of Attention: Higher Empathy Standards, Less Given — AryHHAry · 2026-10-04
- Aidan Maps Jung's Shadow Integration to ML: Brain 'Grokking' Itself via RL — beffjezos · 2026-10-04
- New site collects real-world AI wins — beyond benchmarks and doom — ajratner · 2026-10-04
- Agents could make online competition 10x worse: you're now competing with everyone's agent — signulll · 2026-10-04
- Prediction: a fully AI-generated chart-topping hit within 3-12 months — cnakazawa · 2026-10-04
- 'Human-only' art will become a marketing label, but drawing the AI-assist line is impossible — Afinetheorem · 2026-10-04