It's always interesting to observe someone as brilliant in "one task" - use a common AI phrase - such as Le Cunn being brilliant in AI, and this same person being a complete moron in another. Le Cunn is stating that China and Russia are engaged in corrupting elections of countries. This purely based on American propaganda. Hillary Clinton loses an election because many Americans hate her, then she blames Russia for making her people hate her. And someone as smart as Le Cunn parrots this rubbish. We know that Russia doesn't go around the world making countries follow "Russian democracy". On the other hand, America is all over the world, prescribing "American democracy", through disinformation by its NGOs, and where NGOs fail, through military coups and color revolutions. Interestingly by the way, this is why open source must be encouraged. It is always the concentration of technology in a few hands, that leads to bad actors because that concentration of power makes the one who has it, abuse it. Again America gives us an example of this, when they started terrorizing the world the moment they had a nuclear bomb. Had the Soviet Union not developed its own quickly, the Yankee bully would have committed its evils unbated. We also saw this when the Soviet Union broke down. Left alone as the superpower, America was free to terrorize people in Iraq, Libya, Syria, etc. So Le Cunn is correct, we need to open source all technology, but not to curtail Russia and China, but to ensure that the American bully doesn't ever believe that, it and it 48:59 , has this power.
Not true. They both started studying the back propagation algorithm around 1986. They share the Turing Award for a reason. If anything, David Rumelhart would have deserved being acknowledged here … And it was Hinton’s student Alex Krizhevsky who started the DL hype.
Agreed. He's very smart when it comes to some things, but surprisingly not smart about others. Particularly when it comes to factoring in human motivations and behaviors.
I like when a person of Yann's caliber can give insights over social, politic and other interactions of AI. He may have slight preference bias sometimes but who doesn't really? I enjoyed it very much, thanks a lot Zetta for your initiative of AI Native!
