Mood
Original
11 / Race

AI and human performance

Compare AI with the baseline in each study. Select a point for results and limitations.

AI aheadBeat the human baseline in the study
MatchedComparable to the tested baseline
ContestedDisputed result or rebuttal
Humans aheadHuman baseline performed better
2026
Drag to turn · click a point to read it
3D isn’t available in this browser — every study is in the ledger below.

The ledger

Select a study for its result, baseline and source. Benchmark results do not establish job performance.

Sources and method

Baselines vary: champions, experts, typical participants or human-made methods. Each study names its comparison and caveats. Counts summarise study classifications; individual measures may differ.

Papers, preprints and official competition reports are labelled. Statuses reflect evidence checked 27 October 2026.