About
Predictions about artificial intelligence shape investment, policy, career choices, and public anxiety — yet almost nobody goes back to check whether they came true. Herbert Simon said machines would do any human job by 1985. Geoffrey Hinton said radiologists were finished by 2021. The forecasts made headlines; the outcomes rarely did.
The AI Prediction Index is the missing scoreboard. We collect notable, falsifiable public predictions about AI, wait for their deadlines, and record what actually happened — with sources, resolution notes, and a transparent methodology.
The goal isn’t to embarrass anyone. Forecasting is hard, and the people bold enough to make testable claims deserve more credit than those who only speak in vibes. But a public record of accuracy makes the whole conversation about AI’s future more honest.
Suggest a prediction or a correction
Know a notable, sourceable AI prediction we should track — or think we scored one wrong? We want to hear it. Every entry needs a citation to the original claim, and every verdict is open to challenge on the merits. A public submission channel is coming soon.