Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Yann LeCun, Geoffrey Hinton and Yoshua Bengio win Turing Award

nytimes.com

Post reply on HN
Print view View original

101–110 of 159 posts

Re: Yann LeCun, Geoffrey Hinton and Yoshua Bengio win Turing Award

#101

Post by isoprophlex »

mistrial9 wrote: post #59

as a non-specialist, it seems like several important and distinct areas of inquiry are just lumped together in the celebration of a "winner" for ML Text, language, human chat Image recognition from blurry, multiple views, multiple lighting conditions photos formal patterns with large numbers of variations These are not at all the same, yet the praise seems to want to declare "the best" and "beating competitors" .. wh…

i want to get on the hype machine and holler inaninties while i shoot my revolver into the ceiling.

we must glorify the efforts of a arbitrarily chosen single individual, in the hope of someday becoming that lauded singleton, the one allowed by decree to piss on the heads of those fools below who didn't win the race.

Re: Yann LeCun, Geoffrey Hinton and Yoshua Bengio win Turing Award

#102

Post by return0 »

swalsh wrote: post #4

I have a browser extension that replaces the phrase "Artificial Intelligence" with the phrase "A bunch of if statements". This might have been one of the top results.

You guys need to get over it. Intelligence is not a well defined term. AI is fine do describe something that works "humanlike" even though we don't kno why. I would dare you to give a definition of "AGI"

Re: Yann LeCun, Geoffrey Hinton and Yoshua Bengio win Turing Award

#103

Post by darksaints »

bernardopires wrote: post #99
darksaints wrote: post #98

Earlier quoted context omitted.

You've identified four areas where it excels. Only a few billion more areas to go. And sure, we can now have algorithms that can categorize dogs based on collections of pixels better than humans can. Get back to me when they've figured out double entry accounting, splitting atoms, building bridges and spaceships, or improve upon TCP-IP.

This is a computer science award, nobody is claiming that AI will revolutionize the fields you mention (which are at best loosely related to CS). It doesn't have to be useful literally everywhere to have value.

Yes, people actually are claiming that these new NNet advancements are the key that will unlock AGI, which by its very nature implies that they will be as advanced, or more advanced, than humans in all forms of intelligence. They're lightyears away from that, but now the general public is buying into the hype and thinking it's a decade or two away.

NNets are already a disappointment to nearly any applied researcher that isn't sitting on petabytes of data. It won't be long before the CEOs realize it and put their research funding somewhere else.

Re: Yann LeCun, Geoffrey Hinton and Yoshua Bengio win Turing Award

#106

Post by moultano »

phowon wrote: post #63
nabla9 wrote: post #33

Earlier quoted context omitted.

Schmidhuber's significance in the field is not at the same level. He invented one type of neural network, but the "Canadian mafia" launched the whole Deep Learning revolution with multiple breakthroughs and theoretical understanding. LSTM's are clever and they were bleeding edge up to 2014, but once they were understood better as a bypass mechanism, attention, context vectors and averaging networks and causal convolu…

Putting aside the whole Schmidhuber debate - where are people getting this idea that causal convolutions are anywhere near the prominence of RNNS/LSTMs? As far as I'm aware, causal convolutions were used in WaveNet (and subsequent models) and a small number of NLP applications. Meanwhile, LSTM-based models are used in just about every NLP paper, and at least a baseline in the newer ones more dominated by Transformers…

I think LSTMs have been mostly abandoned. I don't see much new work using them, vs CNNs or transformers.

Re: Yann LeCun, Geoffrey Hinton and Yoshua Bengio win Turing Award

#107

Post by moultano »

li4ick wrote: post #29

I think Schmidhuber should be up there. LSTMs are everywhere, not just NLP. The latest Starcraft 2 bot from Deepmind uses LSTMs extensively in its architecture.

I think we need a few more years to see whether LSTMs retain their importance over time. Awarding it now seems premature.

Re: Yann LeCun, Geoffrey Hinton and Yoshua Bengio win Turing Award

#108

Post by phowon »

moultano wrote: post #106
phowon wrote: post #63

Earlier quoted context omitted.

Putting aside the whole Schmidhuber debate - where are people getting this idea that causal convolutions are anywhere near the prominence of RNNS/LSTMs? As far as I'm aware, causal convolutions were used in WaveNet (and subsequent models) and a small number of NLP applications. Meanwhile, LSTM-based models are used in just about every NLP paper, and at least a baseline in the newer ones more dominated by Transformers…

I think LSTMs have been mostly abandoned. I don't see much new work using them, vs CNNs or transformers.

Aren't causal convolutions basically CNNs with masking?

Re: Yann LeCun, Geoffrey Hinton and Yoshua Bengio win Turing Award

#110

Post by ska »

amelius wrote: post #109

And we still don't know if/how biological brains perform backpropagation. Perhaps we should consider calling NNs "weight networks" instead.

That was pretty much hashed out in the 70s when NN interest got going again. I'm not sure there is much value in recapitulating the discussion....
Post reply on HN