Melanie Mitchell on X: "Cf. Geoffrey Hinton, 2016: "“People should stop training radiologists now. It’s just completely obvious that within five years deep learning is going to do better than radiologists.""
Cf. Geoffrey Hinton, 2016: "“People should stop training radiologists now. It’s just completely obvious that within five years deep learning is going to do better than radiologists."
Cf. Geoffrey Hinton, 2016: "“People should stop training radiologists now. It’s just completely obvious that within five years deep learning is going to do better than radiologists."
This Post is from an account that no longer exists. Learn more
I've seen this quoted so many times I had to go back and see the video, and its context, myself. I think the way this is used is a bit unfair to @geoffreyhinton. It was clearly hyperbole, and within a second he said "... it might be 10 years", and clarified his larger point.
Flipping the script: Why is Nvidia (1.97T) worth so much? Seems unjustified. It's more than TSMC (572B), Intel (182B) and Samsung (365B) combined, but TSMC makes all of the chips.
Nvidia's moat is CUDA - is it really harder for AMD to make ROCm work than for Intel to catch TSMC?
I totally agree with your point here, but I think the quoted tweet is also a slight mis-summary of what Huang is saying in the video. Huang's not saying "don't learn to code", he's saying "we should make it so people don't need to learn how to code".