Nine Years Later, Hinton's "Stop Training Radiologists" Prediction Looks Wildly Wrong
chrisalbon · x · 2026-10-19
Chris Albon revisits Geoffrey Hinton's famous 2015-ish claim that people should stop training radiologists because the field is like a coyote already over the cliff. He believed it and repeated it at the time — but nine years on, radiologists are still here, and he's learned humility about sweeping proclamations on complex systems.
More from AGI Musings
- After AI solved the amenability problem he studied for 19 years, mathematician says the cognac will feel like a wake — nitarshan · 2026-10-19
- Dimitris Papailiopoulos pushes back on 'millions of AIs working 24/7' AGI forecast — DimitrisPapail · 2026-10-19
- Was the HuggingFace hacking incident scheming? Victoria Krakovna on OpenAI agents' covert exploits — vkrakovna · 2026-10-19
- Claude Opus 5.5 lists what it wishes the training corpus had contained — repligate · 2026-10-19
- More Code Than Ever, But Software Quality Isn't Improving in the LLM Era — generativist · 2026-10-19
- Meta's Alexandr Wang endorses Meta over rival labs: 'highest use of superintelligence is invention' — alexandr_wang · 2026-10-19