AI predictions that missed their deadlines · a curated, evidence-screened timeline

In the last ten years, 12 of 29 named forecasters put deadlines on AI that passed unmet — the misses cluster around cars, jobs, agents and AGI

One representative falsified prediction per person: 29 dated, falsifiable AI and automation claims by notable researchers, founders and commentators, stated between 1950 and 2025, every deadline elapsed by August 28, 2026, each with a cited assessment. Twelve of them were made in 2016–2025, and this redesign puts that decade first; the 17 older entries remain below as historical context. The recurring recent error is rarely direction — AI did progress — but confidence, scope, and a short deadline. A miss grades the stated deadline and scope, not motives or a career, and a later arrival would not make the original deadline correct.

filters the historical strip and the table; the last-ten-years timeline is always shown

Last 10 years · 12 people, statements 2016–2025

statement made deadline passed unmet later assessment, if a different year too pessimistic — the direction exception dashed span — low confidence: vague metric or debatable assessment

Colour by topic family: autonomous vehicles jobs and software work AGI and general cognition AI agents AI progress itself

Historical context, 1950–2015 · 17 people, muted for scale

What the timeline cannot say alone

Autonomous vehicles are the densest recent cluster: Carlos Ghosn promised self-driving cars by the end of the decade, Mark Fields promised pedal-free vehicles on streets by 2021, and John Zimmer predicted private ownership would all but end in major U.S. cities by 2025 — none arrived on schedule.qz.com

Rodney Brooks is the decade’s lone too-pessimistic entry: he predicted the popular press would report the deep-learning era ending by 2020; GPT-3 launched that year and the generative-AI boom accelerated instead.forecastfools.com

The windows are getting shorter: Dario Amodei gave three to six months in 2025 for AI writing 90% of code — later public estimates at major companies ran around 25–30% — and Eric Schmidt gave one year for replacing the vast majority of programmers, a window only recently elapsed.sloppish.com

Corporate goals miss too: Marc Benioff’s target of one billion Agentforce agents by the end of 2025 was a stated company goal, and later assessments found a far smaller deployment — graded here on the stated number and date only.dev.to

All 29 people, recent rows first

Person	Claim and outcome	Made → deadline	Topic	Confidence	Source

Confidence and caveats

High confidence means a concrete deadline and a clear outcome; Medium means the interpretation, wording or scope requires care; Low means a vague metric or a debatable assessment, and should not be read as equally conclusive.
Recent claims are harder to grade because terms such as AGI, AI-written code, or programmer replacement lack universal metrics. Mark Zuckerberg’s item was framed as a goal; Sam Altman’s and Andrew Ng’s claims have vague operational thresholds; Marc Benioff’s billion-agent item was also a corporate goal; Dario Amodei’s code percentage can vary by organization; Eric Schmidt’s one-year window is only recently elapsed.
Among the older entries, Alan Turing’s forecast was conditioned on memory capacity, Marvin Minsky later disputed the magazine quote attributed to him, and Shane Legg’s statement was explicitly probabilistic — a modal estimate, not a promise.
The qz.com sources for Carlos Ghosn and Mark Fields may be less accessible; the links are retained as cited in the dataset.
Strict inclusion: a named notable person, a falsifiable AI or automation claim, and a deadline already elapsed by August 28, 2026. One representative prediction per person; repeat offenders appear once. Nothing here alleges deception or incompetence — only the stated deadline and scope are graded.

Curated timeline of 29 falsified AI predictions, exactly one row per named person, statements 1950–2025, deadlines through 2026, all elapsed by the 2026-08-28 cutoff; 12 people made their representative claim in 2016–2025. Each span measures statement year to stated deadline year; a dotted ring marks a later assessment year. Full claim-and-outcome wording is preserved in the table; timeline labels are shortened for space. Curated, not an exhaustive census of mistaken forecasts.