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

The Great A.I. Awakening

nytimes.com

Post reply on HN
Print view View original

71–80 of 147 posts

Re: The Great A.I. Awakening

#71

Post by ryaniscool »

aedron wrote: post #24

And the pattern of overselling A.I. continues. What we have accomplished in this generation is pretty good pattern recognition. Nice, but limited in its usefulness to obvious applications, like translation and image and audio classification. Go and Chess games are rule-constrained enough to be reducible to patterns, so it works for them too, giving a false impression of 'general AI'. What's worse is that this approac…

Yup, in general, people predicting the imminent end of civilization from sentient A.I. lack technical understanding of how current A.I. works.

Yes, it may put you out of a job one day but it won't actively seek to destroy the human race.

Re: The Great A.I. Awakening

#72

Post by YeGoblynQueenne »

earthly10x wrote: post #4

A lot of AI is about vector space these days. https://www.kaggle.com/c/word2vec-nlp-tutorial/forums/t/1234... "That interim also saw dedicated attempts on the part of Google’s competitors to catch up. (As Le told me about his close collaboration with Tomas Mikolov, he kept repeating Mikolov’s name over and over, in an incantatory way that sounded poignant." "Just as the chip-design process was nearly complete, Le and…

>> When you summarize language in a similar way, you essentially produce multidimensional maps of the distances, based on common usage, between one word and every single other word in the language.

The problem with word embeddings, or any distance-based model really, is that language doesn't work that way.

Chomsky has a standard example he uses to make this point: "Instinctively, Eagles that fly swim". He points out that in this phrase, the "instinctively" goes with "to swim" (as in "instinctively, they swim") even though the phrase, and the attachement, mean nothing (the phrase is nonsensical by design).

If the relation was really based on distance, we would expect "instinctively" to attach to "fly". The fact that it doesn't suggests that there is something else that makes us pick the correct association out of all the possible interpretations in that sentence.

Word vectors in their original form also have trouble with homonyms etc "faux amies": for instance, the word "cat"- is it referring to the animal, or to the Linux command? In vector space, there wouldn't be any difference, so the animal would be associated with the symbol ">" and the Linux command with "small" and "furry".

Re: The Great A.I. Awakening

#73

Post by iandanforth »

apsec112 wrote: post #8

Sundar seems very confused here. The idea that we should invent AGI to make "personal digital assistants" is like Hogwarts students inventing time machines so they can take two classes at once. I mean, yes, it would technically enable you to do that, but it doesn't matter because at that point you have way, way bigger problems. If a fleet of alien spaceships arrived in the sky over Manhattan tomorrow, our first react…

Marketing, not confusion.

Re: The Great A.I. Awakening

#74

Post by Quarrelsome »

pixl97 wrote: post #50
Quarrelsome wrote: post #39

Earlier quoted context omitted.

IF AGI is developed. Lets not take is as a certainty. This problem might be beyond our capabilities. Its kinda like stating: When we invent warp-drive. We don't have a proof for warp-drives same as we don't have a proof for building AGI.

>We don't have a proof for warp-drives same as we don't have a proof for building AGI. We don't have a warp drive to copy, no idea how one could even exist, they are not in our known universe. If we want to learn about an AGI, we just look in the mirror.

Fine. Show me the proof and we'll build it.

... except there is no proof. We have no proof but we talk about it like its a given. We talk about "intelligence" as if its an integer value that scales from 0 to infinity. These are ideas that we do not question enough.

I assert that its a possibility that we are incapable of building ourselves beyond reproduction.

Re: The Great A.I. Awakening

#75

Post by iandanforth »

aedron wrote: post #24

And the pattern of overselling A.I. continues. What we have accomplished in this generation is pretty good pattern recognition. Nice, but limited in its usefulness to obvious applications, like translation and image and audio classification. Go and Chess games are rule-constrained enough to be reducible to patterns, so it works for them too, giving a false impression of 'general AI'. What's worse is that this approac…

If we only looked at feed forward conv nets, then you'd be correct. However curriculum learning in environments like Universe and Labs are a critical step toward generality and planning based AIs. Solving catastrophic forgetting and increasing the time interval between behavior and reward are non-trivial steps. I'm not saying it's raining right now, but yeah I'm saying you should buy an umbrella.

Re: The Great A.I. Awakening

#76

Post by varjag »

psyc wrote: post #68
varjag wrote: post #55

Earlier quoted context omitted.

> Compared to every other life form on Earth you are super intelligence. This super intelligence thing has already happened once. You extrapolate the trend from one data point. There is no indication that we have some matyoshka type hierarchy of intelligence. It could well be binary, either intelligent or not. No reason so far to think otherwise. > You can't beat chess in one move because mathematics does not allow i…

> So what superintelligence can do that "ordinary" intelligence in right quantities couldn't? Conceive of answers to this question that we can't. Do you really believe the human imagination captures all of possibility? Why?

The search space of all possibilities is infinite, and hence beyond the grasp of any intelligence. I firmly believe human imagination and all other forms of intelligence are restricted by mathematics and physics of our universe. And as such, the fundamental set of solvable problems remains the same.

Would be nice to hear some specific arguments against instead of Penrose-like quantum handwaving and pet analogies.

Re: The Great A.I. Awakening

#78

Post by Retric »

nopinsight wrote: post #21
apsec112 wrote: post #8

Sundar seems very confused here. The idea that we should invent AGI to make "personal digital assistants" is like Hogwarts students inventing time machines so they can take two classes at once. I mean, yes, it would technically enable you to do that, but it doesn't matter because at that point you have way, way bigger problems. If a fleet of alien spaceships arrived in the sky over Manhattan tomorrow, our first react…

This is why I doubt if efforts like OpenAI will ultimately turn out to be beneficial or catastrophic. I understand their purpose of democratizing the power and benefits of AI technology broadly. However, if their platform can provide a critical component or developmental path to AGI in the future, the unrestrained distribution of the technology is similar to spreading the blueprint of nuclear warheads. It could resul…

Repeat after me, "AI is not Magic."

A super intelligent AI does not change the laws of physics or even get to predict the weather two weeks from now.

Sure, they could learn to be a good hacker but the worlds best hacker can't get into air gaped systems. More importantly, flawless security trumps hacking. If there is nothing to exploit there is no way in, which means a lone AI is much more powerful than lot's of AI's with different goals.

Re: The Great A.I. Awakening

#79

Post by randcraw »

melvinmt wrote: post #6
aaron-lebo wrote: post #2

The use of Hemingway as an example is unfortunate. Something less literary would be a better example maybe?: Close to the western summit there is the dried and frozen carcass of a leopard. No one has explained what the leopard was seeking at that altitude. Near the top of the west there is a dry and frozen dead body of leopard. No one has ever explained what leopard wanted at that altitude. One of those is pretty sti…

> Starting salaries of seven figures are not unheard-of Seven figures? For grad students?

Ruputedly, Andrej Karpathy and perhaps others recently saw offers of that size (presumably including incentives and options).

But the claim has become an pop meme, so who knows.

Re: The Great A.I. Awakening

#80

Post by posterboy »

vanderZwan wrote: post #69
ythn wrote: post #65

Earlier quoted context omitted.

People like believing in science fiction. If it makes them feel better, let them.

The problem is when faulty beliefs lead to faulty expectations, or worse: faulty policies.

Well, can't hurt to raise some awareness long before it's actually problematic. Basically it means to trust any program just as far as trust in the programmer would allow. A bug is a bug, AI or not.
Post reply on HN