Hacker News new | past | comments | ask | show | jobs | submit login

There's a big jump from backprop to what we have now, Hinton mainly does the AI equivalent of fundamental physics not applications.



This is just flat-out wrong. You make it sound like Hinton hasn't done much since his famous back propagation paper, or that he hasn't been intimately involved in productizing some of his research.

Hinton's startup, DNNresearch Inc., which made breakthroughs in machine vision (particularly around identifying objects in images and image classifications), was acquired by Google in 2013, specifically to help with image search (and also, obviously, for the talent of the team). Hinton's cofounders in that startup were Alex Krizhevsky (of AlexNet fame) and Ilya Sutskever, current Chief Scientist at OpenAI.


I aim to make it sound like Hinton isn't in the cutting edge of LLM research - not that he is somehow incapable of it, but rather that anyone who isn't at OpenAI at the moment is probably in the dark. The most recent thing I have seen of him on my feed for example was a paper into the fundamentals of learning (The forward-forward paper), for example.




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: