On: Learning to Love the AI Bubble

[sjg007]

Neural networks add computational depth. So I would disagree with the statement that AI is just "stats rebranded". That's about as useful an analogy as saying that statistics in practice is just applied linear algebra.

[codesushi42 → sjg007]

Define computational depth. Non linearity? Parallelizable? Computational depth sounds like hyperbole.

You're still approaching stats problems with the same methodologies. Your just using NNs as your optimizer.

[sjg007 → codesushi42]

If you are interested, I would suggest reading up on the https://en.wikipedia.org/wiki/Universal_approximation_theore...

[6gvONxR4sf7o → sjg007]

There are theorems like that for polynomials and fourier series and all sorts of other function classes too. They are just as practically relevant (or irrelevant).

[sjg007 → 6gvONxR4sf7o]

Sure... I mean it is matrices all the way down but the claim that AI (e.g. deep learning) is just applied statistics is disingenuous.