Toby Ord calls the 'AI got 100,000x cheaper' chart wrong: capabilities mixed into cost curves
tobyordoxford · x · 2026-09-23
After digging into a widely shared chart, philosopher Toby Ord argues its headline result is incorrect: it claims AI prices fell more than 100,000x from 2021-2026.
His objections:
- Nothing plausibly got 100,000x cheaper; this looks like an aggregate/proxy statistic rather than a real price measure
- The chart appears to merge rising capabilities with falling costs and force them into a pure price-decline figure, which he says isn't possible — so the headline numbers don't make sense
- The x-axis is relative years since a tech's introduction, not real dates; given the claimed high-then-low quarterly declines, the slope should flatten over time, which it doesn't
- Applying the same method to other industries would likely produce similarly steep curves, so it isn't an apples-to-apples comparison
Bottom line: the chart's headline results shouldn't be cited as evidence of collapsing AI costs.
Related event: Scholars Challenge Viral Chart Claiming 100,000x Drop in AI Prices(10 posts)→
More from AGI Musings
- Investor: AI diffusion will be slowest in consumer software — Oura, Strava, Labcorp still have no real AI strategy — MartinGTobias · 2026-11-12
- On AI Progress: When Models Lag on Your Load-Bearing Skill, You Can't See Them Improving — dfrsrchtwts · 2026-11-12
- Slack's David Hoang: agents beat people on coordination, proposes 'boundary agents' — davidhoang · 2026-11-12
- CS student mourns coding flow states he'll never experience in the AI era — chrissyykat · 2026-11-12
- Ford's 370% Turnover Led to $5 Days — But ASI Never Quits — krishnan · 2026-11-12
- Against 'verifiable vs. fuzzy': the real axis is how expensive tasks are to grade — dfrsrchtwts · 2026-11-12