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
@GaryMarcus i read https://arxiv.org/abs/1801.05667 and in particular you list of 10+ primitives that are required to be innate machinery. for some of them, it is not clear to me whether those primitives can all be conceptualized in the language of machine learning. can you clarify...
i don’t think those algorithms have any of them. the point (developed in http://rebooting.ai ) is that we need some new ideas, to take us beyond the existing mechanisms. (CNNs have translational invariance)
ok, but certain ML algs can approximate any turing machine. so you are saying we need something better than turing machines? (i'll read the book and then have more questions)
All DNNs have translational invariance is they are trained on enough data. It's not an intrinsic property. CNNs add only a slight invariance but lose precision. The result is that less data is needed for training. The brain's invariance works even with objects it has never seen.
I would not even call what CNNs do, invariance, as it is defined in connection with the brain. It's more like they add a fuzzy positional property to vectors.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.
