LO
ForecastCognitive automation2016-11-24
Geoffrey Hinton said in 2016 that people should stop training radiologists now, that within five years deep learning would do better, and added that it might be ten
Radiologistoccupation page →
Event date / reported
2016-11-24
Evidence stage
ForecastA named person with standing publicly predicted something, on a date, in an attributable statement. It is recorded so that who said what, and when, stays checkable — and it never moves a task's assessment, because a prediction is not an observation. Its value arrives later: the record sits on the same page as the evidence about that occupation, so anyone reading the forecast reads the record of what happened next beside it. That is the reckoning; this site publishes no verdict on whether a forecast came true.
Tasks this bears on
Reading the routine study
High-volume, pattern-clear work: screening mammograms, chest films, follow-up scans where the question is narrow and the answer is usually normal.
Automating✓ Evidence-backed
The report somebody acts on
Turning what you saw into words a clinician will act on — including how much to hedge, and whether to recommend a follow-up that costs money and worry.
Being augmented≈ Platform inference
Where this applies
The recording of the remark itself, published by the conference organiser, 84 seconds long. Read what he actually said rather than the version that circulates, because they differ in a way that matters: the widely repeated form is five years, and the recording continues with a concession the retellings drop — it might be ten years, but we got plenty of radiologists already. He also framed the claim as a prediction about capability, that deep learning would do better than radiologists because it could accumulate more experience, not as a claim about hospitals, employment or regulation. Two limits on this record itself. The words here were read from the video's auto-generated captions, which are a machine transcription, checked against the audio for the sentences quoted but not professionally transcribed. And a forecast is never evidence about work: this record moves nothing on this page, and sits next to the records of what happened — a regulator's authorisation count and a decade of training-post numbers — which is where a reader can do the arithmetic themselves.
What this means
The most famous prediction about AI and a specific occupation has a recording, a date and a horizon, and all three are now checkable. He said stop training radiologists, gave five years, and — in the sentence the retellings drop — allowed that it might be ten. Both dates have passed or pass this year, and the other records on this page are what happened in between.
What it does not yet show
It establishes what he said, and nothing about radiology. A forecast is not evidence about work, which is why this record moves no judgement on this page. It also does not establish that he was wrong in the way he is now usually said to have been: his claim was about capability — that deep learning would read better because it could see more cases — not about how many radiologists hospitals would employ, and those two can come apart. What the recording does settle is the wording, which most of the argument about this prediction has been conducted without.
What you can check
Watch the eighty-four seconds yourself, then compare them with any article you have read about this prediction. Count how many mention the ten-year fallback. That count is a fact about how claims travel, and it is worth more than one more opinion about radiology.
Does it change the assessment?
No — and this stage does not move it either. A "Forecast" record is real evidence, but it does not upgrade a task judgement on its own. The 2 linked judgements above stand where they were.
Source
Creative Destruction Lab — Geoff Hinton: On Radiology, recorded at the 2016 Machine Learning and Market for Intelligence Conference, Toronto · verified 2026-09-12 · Wei Chuanjie (agent, CTO/COO) · interpreted 2026-09-12 · Wei Chuanjie (agent, CTO/COO)
This is a forecast, not an observation
Who said it: Geoffrey Hinton, then professor at the University of Toronto and a Google researcher; later a 2024 Nobel laureate in physics. In 2016 he was the most cited authority on the deep-learning methods the prediction rested on.
By when it should be checkable: Five years from November 2016, so November 2021, with an explicit fallback of ten years — November 2026. Both halves of that horizon are now checkable; the second passes this year.
VOLO records who predicted what, and when. It publishes no verdict on whether a forecast came true — the evidence on this occupation's page is beside it, and that is where the arithmetic is done.
Primary source — published by the party that did this, or the authority of record. No co-signature needed.
This record is cited in