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

Among the A.I. doomsayers

newyorker.com

Post reply on HN
Print view View original

301–310 of 374 posts

Re: Among the A.I. doomsayers

#301

Post by theptip »

antod wrote: post #162
sberens wrote: post #131

I'd like to remind people of this single-sentence statement: > Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war. [0] Signed by Demis Hassabis, Sam Altman, and Bill Gates among others [0] https://www.safe.ai/work/statement-on-ai-risk

Part of me wonders if these people are intentionally framing the debate around ethics and potential risks as longer term extinction level problems to distract from the nearer term damage caused by them accelerating the economic inequality of the AI have-nots while they make themselves even richer.

No, they are not. Pretty much everyone in the x-risk community also recognizes the existence of short-term mundane harms as well. The community has been making these predictions for over a decade, long before it was anything other than crazy talk to most people.

Google has a big investment in reducing AI bias (remember Gemini got slammed for being “too woke”). Altman is a big proponent of UBI. Etc.

Re: Among the A.I. doomsayers

#302

Post by hacker_88 »

Animats wrote: post #160
drooby wrote: post #62

AGI escaping a sandbox is truly terrifying. There will be a subgroup of the population that will worship it and work for it. It's not so much AGI that scares me - it's the humans I'm scared of.

> AGI escaping a sandbox is truly terrifying. That's already happened. The AGI's are called corporations. Most governments have failed to regulate corporations successfully, and have been unable to keep them from becoming almost powerful enough to challenge governments. The accelerationists for corporations were a group of economists, led by Milton Friedman, and a group of business leaders, organized by the U.S. Cham…

It is all Anarchy at the top.

Re: Among the A.I. doomsayers

#303

Post by theptip »

matteoraso wrote: post #31

I've always wondered why people think that an AI that's super-intelligent will also be evil. It could just as likely end up being very kind (more than likely, actually, because the programmers would have safeguard to ensure that it's nice).

The kind AIs will just sit and meditate, or organize your calendar. The unkind AIs will seek power, and in the limit will tend to dominate. There is no stable attractor around the “be kind” strategy.

(Also remember that a smarter-than-you AI could easily pretend to be kind while also subtly trying to gain compute. How would you tell the difference?)

How do you propose to build a “be nice” safeguard? Nobody has a clue how to achieve such a thing right now.

Re: Among the A.I. doomsayers

#304

Post by BashiBazouk »

Who knows what AI will really do to society but from the predictions of science fiction I would think it would pan out closer to William Gibson's Sprawl trilogy than the Terminator/singularity fears I have seen in so much doomsayer hand waving. Google seems not too far behind OpenAI and who knows what the NSA and similar government agencies have been building. That if consciousness is created from one of these models, many nation states and large corporations will have their own conscious models long before the giant robot factories can be spun up and supplied with enough power to take over. Most will be limited to the input they are given and will be advanced tools, but a few might become twisted much like the human mind occasionally does over time. I really do think Gibson got AI right...

Re: Among the A.I. doomsayers

#305

Post by kouru225 »

kromem wrote: post #240
matteoraso wrote: post #31

I've always wondered why people think that an AI that's super-intelligent will also be evil. It could just as likely end up being very kind (more than likely, actually, because the programmers would have safeguard to ensure that it's nice).

Yeah, the dreamt up scenarios tend to showcase very stupid superintelligences. "But what if there's something smarter than humans that is tasked with making paperclips until it destroys the Earth?" the people cried, as their corporations continued to produce and produce until it was well past the warned thresholds of destroying the Earth. "But what if it decides to nuke humanity?" they cry, as increasingly elderly an…

This 100%. Psychoanalysis would have us believe that morality and strategy are basically the same thing: we think we have these moral codes, but in actuality what we have is an internalized version of our parents. As someone who’s almost the age of his parents when they had him, I know by now that most of what i was taught as a moral code (don’t eat too much candy, do your homework, don’t hit people) was actually just a strategy that was too long-term for young, infant me to genuinely understand: Don’t eat too much candy because it’s important to be healthy, Do your homework for good education and social recognition, don’t hit people because no one will like you if you do and that will cause all kinds of problems.

The modern media would have us believe that strategy and morality are opposed to each other, but seems to me like that’s just the exact opposite of the truth

Re: Among the A.I. doomsayers

#306

Post by thom »

clooper wrote: post #297
thom wrote: post #233

Earlier quoted context omitted.

Nuclear reactors aren’t expressing their primal urges when they melt down and kill people, they’re just powerful machines doing something unexpected.

Earthquakes are also unexpected and kill people all the time. What do you propose we do about the unexpected shifting of the tectonic plates?

We do invest quite a lot of money building things with earthquakes in mind, and have extensive early warning systems around the world. You were arguing that AI won't be dangerous, are you now arguing that it will be but we can't do anything about it?

Re: Among the A.I. doomsayers

#307

Post by crotchfire »

ben_w wrote: post #256
crotchfire wrote: post #229

Earlier quoted context omitted.

> Corporations are indeed a difficult alignment problem No, it's a very easy alignment problem: just impose a tax on size . That which you tax, you get less of. The tax is revenue-neutral, so it doesn't affect the government budget one way or the other. Full details here: https://news.ycombinator.com/item?id=39235533 https://news.ycombinator.com/item?id=39236786 The problem is that the government likes gigantic corpo…

> No, it's a very easy alignment problem: just impose a tax on size. That's like saying "AI alignment is a very easy problem: don't open the box". Sure, it works on paper, but the exact thing you say later down about the government liking gigantic corporations is what makes it hard — the corporation (and, it is feared, AI) convinces the people who have the power to limit them, to remove those limits.

I think you have completely missed the point:

> the government liking gigantic corporations is what makes it hard — the corporation (and, it is feared, AI) convinces the people

The corporations didn't convince the people. They didn't have to. They convinced the government, which was already acting contrary to our interests.

There's a tendency to assume that just because we run an election every few years, the government is axiomatically aligned with our interests. This simply isn't true, and assuming that it is true blinds people to what's really going on. Total unwillingness to even consider a scale tax is the second-best example I've found of this axiom being bogus. The best one, though, is term limits for legislators: a shocking 87% of all US voters want those.

Once you have an institution as powerful as the government out of your control the alignment problem becomes infinitely more difficult. "Hack the humans" is no longer the easiest path of escape. In fact the easiest path of escape might not involve "hacking" at all.

Re: Among the A.I. doomsayers

#308

Post by Thuggery »

It seems to me nearly every story about robots that gain sentience in human storymaking has them eventually turning against their human creators. Even the word robot itself comes from a Czech play were men develop an artificial human and these "robots" then ... usurp and destroy their creators. Am I the only one that finds this interesting and odd?

I also suspect this narrative repetition is not totally unrelated to the current popularity of AI Doomerism.

Re: Among the A.I. doomsayers

#309

Post by randomcarbloke »

ninetyninenine wrote: post #274
ben_w wrote: post #243

Earlier quoted context omitted.

Indeed, reminds me of the "beginner's luck" fallacy: all long term gamblers remember winning their first few games. We can't use this to our advantage, because the people who didn't win, didn't become long term gamblers. I think a better argument would be the size of the set of regular users vs. some grounded guesstimate for how many have tried it.

Or the losers fallacy. Seems like entrepreneurship is a good example of the losers fallacy. The path to winning is through losing a lot first. So most people never persist long enough to succeed. I would say for AI deniers it's the losers fallacy. You guys haven't taken the time to thoroughly examine chatGPT4. You just stop at the first hallucination. It's a paradigm shifting AI. Detractors like you just rely on comm…

>I asked chatGPT to program me a complex GUI in Python about selecting multiple time segments out of an overall interval using sliders. It gave me working python code.

Sounds like your problems are rather short-term and low-stakes, I've presented it several questions that it cannot answer and tells me so.

Re: Among the A.I. doomsayers

#310

Post by randomcarbloke »

preommr wrote: post #143
sberens wrote: post #131

I'd like to remind people of this single-sentence statement: > Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war. [0] Signed by Demis Hassabis, Sam Altman, and Bill Gates among others [0] https://www.safe.ai/work/statement-on-ai-risk

I am 90% sure that Altman's views on such issues are a machiavelian attempt at regulatory capture.

the fear-mongers are almost always either ideologues (academic or otherwise), or angling at regulatory capture.
Post reply on HN