Overview
AGI tracked as dimension claims.
This tracker asks what needs to be true for AI to count as AGI, then breaks each dimension into lower-level questions backed by curated evidence and confidence judgments.
Composite progress
The average progress across all 9 tracked dimensions, calibrated against human-expert performance.
2
Questions met
17
In progress
18
Not met
Index dial
Data collection is ongoing — scores are provisional and not yet finalized.
Methodology
The public tracker is generated from a maintained workbook. The Questions sheet holds the dimension and question structure, and the Evidence sheet holds the published evidence entries linked to those questions.
Tracker board
Implementation dimensions
Each card represents a high-level AGI statement. Open a card to inspect the current rationale, question status, and published evidence.
Recent updates
Latest dimension changes
AI can correctly understand, reason through, and plan difficult tasks
Updated Apr 10, 2026
AI can adapt and generalize beyond the exact examples it has seen
Updated Apr 10, 2026
AI can stay grounded in facts and evidence and be honest about uncertainty
Updated Apr 10, 2026
AI can manage uncertainty and improve its own answers
Updated Apr 10, 2026
AI can interact with people clearly, helpfully, and with social awareness
Updated Apr 10, 2026
Community
Suggestion queue
Separate factual accuracy from factual honesty more explicitly
The new evidence dimension covers both, but a future revision could make bluffing-versus-being-wrong a more visible distinction in the public wording.
A. Chen | Mar 18, 2026
Add a stronger threshold for long-horizon drift resistance
Repeated-use stability may deserve its own more prominent card or a more demanding question set under the reliability dimension.
J. Morales | Mar 20, 2026
Expose the hidden capability coverage map more clearly
A lightweight legend showing which deeper dimensions each card covers could help expert users audit completeness without confusing general audiences.
R. Singh | Mar 14, 2026