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

Yann LeCun: AI one-percenters seizing power forever is real doomsday scenario

businessinsider.com

Post reply on HN
Print view View original

81–90 of 873 posts

Re: Yann LeCun: AI one-percenters seizing power forever is real doomsday scenario

#81

Post by logicchains »

thomasahle wrote: post #72
aeternum wrote: post #63

Earlier quoted context omitted.

Arguably, this has already happened at least twice in the past. One with the discovery of the Black-Scholes model, and arguably again with the advent of HFT (and the ability to basically front-run trades). The interesting thing about markets is you generally can only make money when things are mispriced. This limits the total potential gain even for actors with perfect information. Suppose we did have an AI model tha…

> The interesting thing about markets is you generally can only make money when things are mispriced. Why would you think things are not terribly mispriced right now? If only we were a lot smarter we'd know how.

>If only we were a lot smarter we'd know how.

That's not true. Pricing and predicting in the real economy depends on information; for a given amount of information, there are vastly diminishing returns for further intelligence. Because intelligence only allows predicting a bit further in future, as the complexity of predicting the future increases exponentially with lookahead distance; it's O(e^n). This is why hedge funds pay for things like satellite footage of oil tankers to predict changes in supply.

Re: Yann LeCun: AI one-percenters seizing power forever is real doomsday scenario

#82

Post by nostromo »

ghshephard wrote: post #68
PeterStuer wrote: post #48

People focus on LLM's and diffusion models because they are so omnipresent now. But an AI for stock picking and prediction that would be realy next level and outcompete the current ones cosistently would siphon of so much wealth it would basically own society if the operators were clever enough not to get so gready short term that the system would litterally collapse overnight.

I'm not following that argument - such a system might siphon off the trading wealth, but presumably that's miniscule compared to the equity wealth that is assigned to the owners of the companies and the wealth that they create?

Yeah, I honestly think this would be a good thing. ML already exists as market makers. It’s a big-ish industry maybe, but not massive. It probably should be as automated as possible which would reduce the risk free profit to be made to next to nothing.

Re: Yann LeCun: AI one-percenters seizing power forever is real doomsday scenario

#83

Post by brutusborn »

homefree wrote: post #70

I think Yann is probably wrong. He refuses to engage earnestly with the “doomer” arguments. The same type of motivated reasoning could also be attributed to himself and Meta’s financial goals - it’s not a persuasive framing. The attempts I’ve seen from him to discuss the issue that aren’t just name calling are things like saying he knows smart people and they aren’t president - or even that his cat is pretty smart an…

Who do you find persuasive and what material should I read/watch to understand their POV? So far I’ve read a bunch of lesswrong posts and listened to some Eliezer talks but still can’t understand the basis of their arguments or when I do, they seem very vague.

The only suggestion that makes sense to me is from the FEP crowd. Essentially if someone sets up at AI with an autopoietic mechanism then it would be able to take actions that increase its own likelihood of survival instead of humans. But there don’t seem to be any incentives for a big player to dedicate resources to this, so it doesn’t seem very likely. What am I missing?

Re: Yann LeCun: AI one-percenters seizing power forever is real doomsday scenario

#84

Post by sgu999 »

shreyshnaccount wrote: post #59
shreyshnaccount wrote: post #58

Earlier quoted context omitted.

You've got the causality the other way round. We are in an apparent state of abundance partly because of exploitative practices

(shit, not you, the gp)

I'll reply anyway ;D

We didn't manage to have an apparent state of abundance with slavery, which is arguably the finest of our exploitive practices. We do manage to have a state of abundance because of the ubiquity of cheap sources of energy, oil in particular.

I'm deeply convinced that current exploitive practices are a consequence of inequalities, no resources scarcity.

Re: Yann LeCun: AI one-percenters seizing power forever is real doomsday scenario

#85

Post by PeterStuer »

samus wrote: post #73
PeterStuer wrote: post #61

Earlier quoted context omitted.

If I watch my kids and you watch yours, our contribution to GDP is 0$. If I babysit your kids and you pay me 1.000.000$, and you watch my kids and I pay you 1.000.000$, we just created 2.000.000$ in GDP. Were we not in abundance the day before but now we are? What changed?

This device only works if these are two completely equal services served by two equally capable persons with the same investment of time and resources. And it assumes that we mutually trust each other to actually do the thing. Also, bringing money into the game makes the responsibility stronger. Especially if those two persons are not living in a vacuum, but have other bills to pay [edit: i.e., to make other people d…

One could argue GDP measures the degree of financialization of a society more than it measures the presence or absense of abundance.

Re: Yann LeCun: AI one-percenters seizing power forever is real doomsday scenario

#86

Post by matjus »

The most notable thing about AI right now is that it's the new widget. The economy wanted it, the zeitgeist wanted it, and for that purpose no more development is necessary. It's already reshaped McKinsey's advice, Microsoft's operating system and the public's trust in the exceptionalism of art creation.

take Auto-Tune: before it blew up, one uncritically accepted that a great sounding vocal performance was simply that. The mere existence of the tool broke a covenant with the public -- artists could assume good faith on behalf of listeners once, but no more.

Similarly, AI's chief effects are likely to be cultural first, and material second. They've already broken the "spell" of the creator. Seemingly overnight, a solution to modern malaise (choice fatigue, lack of education, suspicion of authority) has colonized the moment.

In this sense, one-percenters "seizing power forever" really have found the best possible time to do so -- I can't recall a time where the general populace was this vulnerable, ill-informed, traumatized and submissive.

That the underlying tech barely works (maybe that will change, but I predict it won't) doesn't really matter.

I don't generally support overly regulatory regimes, but in this case I think existing thinking around monopoly (particularly as it affects the psyche of the aspiring American) is sufficient to indicate something needs to happen

Re: Yann LeCun: AI one-percenters seizing power forever is real doomsday scenario

#87

Post by homefree »

brutusborn wrote: post #83
homefree wrote: post #70

I think Yann is probably wrong. He refuses to engage earnestly with the “doomer” arguments. The same type of motivated reasoning could also be attributed to himself and Meta’s financial goals - it’s not a persuasive framing. The attempts I’ve seen from him to discuss the issue that aren’t just name calling are things like saying he knows smart people and they aren’t president - or even that his cat is pretty smart an…

Who do you find persuasive and what material should I read/watch to understand their POV? So far I’ve read a bunch of lesswrong posts and listened to some Eliezer talks but still can’t understand the basis of their arguments or when I do, they seem very vague. The only suggestion that makes sense to me is from the FEP crowd. Essentially if someone sets up at AI with an autopoietic mechanism then it would be able to t…

There’s a decent podcast interview between Sam Harris and Eliezer Yudkowsky (on Sam’s pod), I think that’s a decent introduction and they break down the ideas in a way that’s more approachable for someone curious about it.

For my personal quick summary I have earlier comments: https://news.ycombinator.com/item?id=36104090

Re: Yann LeCun: AI one-percenters seizing power forever is real doomsday scenario

#88

Post by logicchains »

brutusborn wrote: post #83
homefree wrote: post #70

I think Yann is probably wrong. He refuses to engage earnestly with the “doomer” arguments. The same type of motivated reasoning could also be attributed to himself and Meta’s financial goals - it’s not a persuasive framing. The attempts I’ve seen from him to discuss the issue that aren’t just name calling are things like saying he knows smart people and they aren’t president - or even that his cat is pretty smart an…

Who do you find persuasive and what material should I read/watch to understand their POV? So far I’ve read a bunch of lesswrong posts and listened to some Eliezer talks but still can’t understand the basis of their arguments or when I do, they seem very vague. The only suggestion that makes sense to me is from the FEP crowd. Essentially if someone sets up at AI with an autopoietic mechanism then it would be able to t…

>What am I missing?

You're missing that intelligence is like magic, and enough of it will allow AI to break the laws of physics, mathematics and computation, apparently.

Re: Yann LeCun: AI one-percenters seizing power forever is real doomsday scenario

#89

Post by brutusborn »

PeterStuer wrote: post #48

People focus on LLM's and diffusion models because they are so omnipresent now. But an AI for stock picking and prediction that would be realy next level and outcompete the current ones cosistently would siphon of so much wealth it would basically own society if the operators were clever enough not to get so gready short term that the system would litterally collapse overnight.

I have wondered if this is the risk that Elon is worried about. But he can’t discuss it explicitly because he is working on such a system. If he communicates its existence, then he reveals a huge attack surface and makes his life (and his teams) much more difficult.

So he is limited to publically saying that AI is dangerous but not revealing the true failure mode.

Re: Yann LeCun: AI one-percenters seizing power forever is real doomsday scenario

#90

Post by bafe »

matjus wrote: post #86

The most notable thing about AI right now is that it's the new widget. The economy wanted it, the zeitgeist wanted it, and for that purpose no more development is necessary. It's already reshaped McKinsey's advice, Microsoft's operating system and the public's trust in the exceptionalism of art creation. take Auto-Tune: before it blew up, one uncritically accepted that a great sounding vocal performance was simply th…

I think a (likely beneficial) long-term side effect will be to teach most people to be more skeptical of any artifacts, be it images, audio, video of text. Unfortunately I assume there will be a learning phase where we will pay dearly for this through massive manipulation of public opinion
Post reply on HN