You can add location information to your Tweets, such as your city or precise location, from the web and via third-party applications. You always have the option to delete your Tweet location history. Learn more
Deep learning is useful because it enables us to create programs that we could not otherwise code by hand. But the space of programs you can learn via deep learning models is a minuscule slice of the space of programs that we may be interested in.
Most interesting processes cannot be learned by a stack of affine transforms (plus squashing functions) mapping one vector space to another point-by-point. Even given large amounts of data.
Perhaps more importantly, although human engineers can cover a dramatically larger section of the space of interesting programs (compared to deep learning models), many useful programs remain inaccessible even to humans -- we cannot develop them by hand.
It will be the job of future AI systems to give us access to these programs.
What do you mean by 'interesting' @fchollet ?
Yes. I'm also interested to know a few examples of what you mean by interesting.
I imagine that would be an interesting conjecture to prove. I think it's easier to prove it wrong by the conjecture that neural nets trained using backprop is Turing complete. NNs are, but I've not seen a proof that includes the learning algorithm. Also what's max(t)?
Can you expand on the "cannot" here? Is this a claim about practicality, or theory?
It is about practicality, given the weak ability of humans to write "big" algorithms. And you can theorize that with Kolmogorov complexity. Interesting algorithms may have a high K. complexity, e.g. human intelligence may be few GB of incompressible code.
Isn't every algorithm, if seen as a black box, an affine transform between two different vector spaces?
Better tools for integrating structured and unstructured knowledge and reasoning in combination with backprop learning of said affine transforms seems like a promising path - lot of good work going on in the area of old AI meets new.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.
