Generative AI Fundamentals Map
Map/Models and thresholds/Wave 7

AGI: top-expert-level AI

AI that matches top human experts across most cognitive work. There is no shared definition: AI Futures calls it top-expert-dominating AI, Amodei a country of geniuses in a datacentre, and Hendrycks et al. set a lower bar of a well-educated adult.

stageSpeculative Open on the map →

Facts

Lane
Models and thresholds
Wave
7 of 12 (build order, not a date)
Needs
18 boxes all the way back through hard links
Unlocks
2 boxes downstream through hard links
Leverage
0.1 points · stands in the way of 1 of 53 stages · rank 50 of 57
Cluster
Agents at work
When
Sourced forecast: 2029 to 2033, tail to 2047

Evidence as of October 2026

AI Futures’ forecasters put top-expert-dominating AI at medians from November 2028 to July 2032. Metaculus forecasters put AGI at 25% by 2029 and 50% by 2033. On the lower Hendrycks bar GPT-5 scores 57%, with long-term memory the weakest domain.

  1. 01AI Futures: Q2.5 2026 timelines update (Aug 2026)
  2. 02Dario Amodei: Machines of Loving Grace
  3. 0380,000 Hours: Shrinking AGI timelines, a review of expert forecasts
  4. 04A Definition of AGI (Hendrycks et al.)

Watch
Hendrycks AGI score (GPT-5: 57%), especially the long-term memory domain.

Timeline note
Metaculus forecasters: 25% by 2029 and 50% by 2033. The 2023 survey of AI researchers put the median at 2047.

  1. 0180,000 Hours: Shrinking AGI timelines, a review of expert forecasts

Around it

Needs · 7
Unlocks · 2

Loops and off-ramps

Openings researchers' judgement, October 2026 scouting

Startup room 37/100 · rank 80 of 124 Research room 32/100 · rank 114 of 124
Crowding5 Lab overlap5 Small-team fit1 Buyer today5 Research openness3
Depth
light
Players
OpenAI; Anthropic; Google DeepMind; Meta MSL; xAI; Qwen; DeepSeek
Wedge
No small-team wedge; lab territory (Remote Labor Index best 15.8%). none
  1. 01CAIS: digital labor automation