What percentage of US GDP will be directly attributable to AI systems by 2030?
model spread ±8.15%
90% interval
width ±13.70%
The ensemble median sits at 9.50% with a 90% interval of 4.80% to 18.50%. The four medians sit within ±8.15% of each other, but the models are diverging on the shape of the distribution, with a 90% band ±13.70% wide — the disagreement lives in the tails and the drivers below, not the headline number. The horizon resolves Dec 31, 2030, 1641 days out.
Confidence reflects cross-model spread relative to scale. The p5—p95 band is the collective 90% interval. Models shows whether the four agree on shape.
What percentage of US GDP will be directly attributable to AI systems by 2030?
How the ensemble median and uncertainty band have moved across the polling window.
Expanded by default. Driver-level distributions explain movement in the anchor distribution.
Driver distributions explain movement in the anchor distribution. They do not mechanically generate it.
Expanded by default. Compare how each model views every driver.
Compare every driver across the four-model ensemble. Use this surface to find disagreement, weak confidence, and drivers worth opening.
Each model forecasts the quintile distribution independently. The Crene aggregate is the median of each percentile across these four models.
Missing percentile values in response
Resolved against BEA satellite accounts for AI economic contribution, or if unavailable, the consensus of McKinsey Global Institute, Stanford HAI, and Brookings estimates of AI direct contribution to US GDP.