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
I get that this is another "machine learning discovers that evolution is an optimization process" tweet, but I also feels like it misses the point 
A 30 yo has "hundreds millions" years of experience and their learning abilities aren't half as impressive as a 4 yo.https://twitter.com/karpathy/status/1206064174129930240 …
Hmm... I take Andrej's tweet as a reminder that the "one shot learning" we talk about when trying to compare to biology in fact has a lot of learning behind it. It does seem that it's a bit overlooked (in ML). Then again I barely know what I'm talking about here, honestly.
Oh, I do get his point. It's very old news even if ML people seem to rediscover it every other month. It's just... his tweet is subtly wrong in enough ways to make me unsubtly annoyed 
I feel a bit dense but can you expand on what he's getting wrong? I do want to know!
4yo child is one thing but we need to explain how Megaphragma mymaripenne can fly and navigate with a brain of only 7400 neurons. Each neuron must be doing much more (1000x) than our Perceptron model explains.pic.twitter.com/xNPq0RAgPj
That's pretty amazing. Did not know that!
Agree completely with @sina_lana. This kind of confusion is not expected from someone so eminent and smart like Andrej and it is harmful especially in the current scenario where a lot of people in ML are easily swayed by celebrity researchers.
Also, the encoding of millions of years of experience relates to something Anthony Zador calls Genomic Bottleneck that is important to explain behaviors of precocial species. David Ha has done some really interesting research starting in this directionhttps://weightagnostic.github.io/
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.
