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

Statement on AI Risk

safe.ai

Post reply on HN
Print view View original

491–500 of 964 posts

Re: Statement on AI Risk

#491

Post by arp242 »

kypro wrote: post #432
arp242 wrote: post #318

Earlier quoted context omitted.

This page talks about "extinction from AI". I'm sorry, but I think that's a complete non-issue for the foreseeable future. I just don't see how that will happen beyond spectacular science fiction scenarios that are just not going to happen. If that makes me a Russian bot then, well, хорошо! The risks from AI will be banal and boring. Spam, blogspam, fake articles, fake pictures, what-have-you. Those things are an iss…

Apologies, the Russian bot comment was more me venting frustration at the prevalence of low-effort response like yours (sorry) to those who try to raise concerns about AI safety. I do agree with you that extinction from AI isn't likely to be an issue this decade. However, I would note that it's difficult to predict what the rate of change is likely to be once you have scalable general intelligence. I can't speak for…

> low-effort response like yours

That wasn't my comment; I agree it was low-effort and I never would have posted it myself. I don't think they're a Russian bot though.

As for the rest: I just don't see any way feasible way AI can pose any serious danger unless we start connecting it to things like nuclear weapons, automated tanks, stuff like that. The solution to that is simple and obvious: don't do that. Even if an AI were to start behaving maliciously the solution would be simple: pull the plug, quite literally (or stop the power plants, cut the power lines, whatever). I feel people have been overthinking all of this far too much.

I also don't think climate change is an extension-level threat; clearly we will survive as a species. It's just a far more pressing and immediate economic and humanitarian problem.

Re: Statement on AI Risk

#492

Post by EamonnMR »

deegles wrote: post #480

Serious question... where can I read the best summaries of the arguments in favor of "AGI will destroy humanity"? and also arguments against? I'm not convinced we can predict how it will behave.

Stuart Russell's book Human Compatible articulates it well. He agrees that we can't predict how it would behave, and that's what he's worried about.

Re: Statement on AI Risk

#493

Post by goolulusaurs »

adverbly wrote: post #417
goolulusaurs wrote: post #388

Throughout history there have been hundreds, if not thousands of examples of people and groups who thought the end of the world was imminent. So far, 100% of those people have been wrong. The prior should be that the people who believe in AI doomsday scenarios are wrong also, unless and until there is very strong evidence to the contrary. Vague theoretical arguments are not sufficient, as there are many organizations…

Fun! Let me try one: Throughout history there have been millions, if not billions of examples of lifeforms. So far, 100% of those which are as intelligent as humans have dominated the planet. The prior should be that the people who believe AI will come to dominate the planet are right, unless and until there is very strong evidence to the contrary. Or... those are both wrong because they're both massive oversimplific…

> So far, 100% of those which are as intelligent as humans have dominated the planet.

This is a much more subjective claim than whether or not the world has ended. By count and biomass there are far more insects and bacteria than there are humans. It's a false equivalence, and you are trying to make my argument look wrong by comparing it to an incorrect argument that is superficially similar.

Re: Statement on AI Risk

#494

Post by randomdata »

mitthrowaway2 wrote: post #476
randomdata wrote: post #459

Earlier quoted context omitted.

Exactly. While there is an argument to be made that people are the real danger, that is beyond the discussion taking place. It has already been accepted, for the sake of discussion, that the nuclear missile is the danger, not the math which developed the missile, nor the people who thought it was a good idea to use a missile. Applying AI to the missile still means the missile is the danger. Any use of AI in the scope…

You said that "AI is an application of math. It is not clear how that can be used directly to harm a body." I was trying to illustrate the case that if humans can develop harmful things, like nuclear weapons, then an AI that is as smart as a human can presumably develop similarly harmful things. If the point you are trying to make is that an AI which secretly creates and deploys nuclear, biological, or chemical weapo…

> I was trying to illustrate the case that if humans can develop harmful things, like nuclear weapons, then an AI that is as smart as a human can presumably develop similarly harmful things.

For the sake of discussion, it was established even before I arrived that those developed things are the danger, not that which creates/uses the things which are dangerous. What is to be gained by ignoring all of that context?

> I really don't know what to say to that. Sure, I guess?

Nothing, perhaps? It is not exactly something that is worthy of much discussion. If you are desperate for a fake internet battle, perhaps you can fight with earlier commenters about whether it is nuclear missiles that are dangerous or if it is the people who have created/have nuclear missiles are dangerous? But I have no interest. I cannot think of anything more boring.

Re: Statement on AI Risk

#495

Post by EamonnMR »

This kind of statement rings hollow as long as they keep building the thing. If they really believed it was a species killing asteroid of a cultural project shouldn't they, I dunno, stop contributing materially to it? Nuclear physicists famously stopped publishing during a critical period...

Re: Statement on AI Risk

#496

Post by toth »

yellow_postit wrote: post #484
computerphage wrote: post #387

Earlier quoted context omitted.

Geoffrey Hinton quit google.

It’s hard not to look at his departure through a cynical lens. He’s not been supportive of other critics, both from and outside of Google. He also wants to use his history to (rightfully) claim expertise and power but not to offer solutions.

I disagree. My read on him is that until very recently (i.e., possibly when GPT4 came out) he didn't take x-risks concerns seriously, or at least assumed we were still many decades away from the point where we need to worry about them.

But the abilities of the latest crop of LLMs changed his mind. And he very publicly admitted he had been wrong, which should be applauded, even if you think it took him far too long.

By quitting and saying it was because of his worries he sent a strong message. I agree it is unlikely he'll make any contributions to technical alignment, but just having such an eminent figure publicly take these issues seriously can have a strong impact.

Re: Statement on AI Risk

#497

Post by staunton »

a_bonobo wrote: post #74
sebzim4500 wrote: post #61

Earlier quoted context omitted.

Now it is, when climate scientists were first sounding the horn they got the same response that these people are getting now. For example: "This signatory might have alterior motives, so we can disregard the whole statement" "We haven't actually seen a superintelligent AI/manmade climate change due to CO2 yet, so what's the big deal?" "Sure maybe it's a problem, but what's your solution? Best to ignore it" "Let's foc…

We have had tangible proof for climate change for more than 80 years; predictions from 1896, with good data from the 1960s. What you are falling for are fossil industry talking points. We have had not any proof that AI will pose a threat as OpenAI and OP's link outline; nor will we have any similar proof any time soon.

In retrospect, you can find tangible proof from way back for anything that gets accepted as true. The comparison was with how climate change was discussed in the public sphere. However prominent the fossil fuel companies' influence on public discourse was at the time, the issues were not taken seriously (and sill aren't by very many). The industry's attempts to exert influence at the time were also obviously not widely known.

Rather than looking for similarities, I find the differences between the public discussions (about AI safety / climate change) quite striking. Rather than stonewall and distract, the companies involved are being proactive and letting the discussion happen. Of course, their motivation is some combination of attampted regulatory capture, virtue signaling and genuine concern, the ratios of which I won't presume to guess. Nevertheless, this is playing out completely differently so for from e.g. tobacco, human cloning, CFCs or oil.

Re: Statement on AI Risk

#499

Post by pixl97 »

Tumblewood wrote: post #295
silverlake wrote: post #131

Earlier quoted context omitted.

This doesn’t work either. The consequence of extinction is infinity (to humans). Likelihood * infinity = infinity. So by hand-waving at a catastrophic sci-fi scenario they can demand we heed their demands, whatever that is.

This line of reasoning refutes pie-in-the-sky doomsday narratives that are extremely unlikely, but the case for AI extinction risk justifies a relatively high likelihood of extinction. Maybe a 0.0000000001% chance is worth ignoring but that's not what we're dealing with. See this survey for the probabilities cutting-edge AI researchers actually put on existential risk: https://aiimpacts.org/2022-expert-survey-on-prog…

Existential risk is one of those problems that nearly impossible to measure in most cases.

In some cases like asteroids, you can look at the frequency of events, and if you manage to push a big one of of your path then you can say the system worked.

But is much more difficult to measure a system that didn't rise up and murder everyone. Kind of like measuring a bio-lab with a virus that could kill everyone. You can measure every day it didn't escape and say that's a win, but tells you nothing about tomorrow and what could change with confinement.

Intelligence represents one of those problems. AI isn't going to rise up tomorrow and kill us, but every day after that the outlook gets a little fuzzier. We are going to keep expanding intelligence infrastructure. That infrastructure is going to get faster. Also our algorithms are going to get better and faster. One of the 'bad' scenarios I could envision is that over the next decade our hardware keeps getting more capable, but our software does not. Then suddenly we develop a software breakthrough that makes the AI 100-1000x more efficient. Like lighting a fire in dry grass, there is the potential risk for an intelligence explosion. When you develop the capability, you are now playing firefighter forever to ensure you control the environment.

Re: Statement on AI Risk

#500

Post by somenameforme »

kypro wrote: post #307
tasubotadas wrote: post #236

If only we would fight for the real issues like climate change, instead of fantasies that would be great.

What's the point in dismissing the need for AI safety? Are you guys Russian bots, or do you genuinely see no reason to worry about AI safety? But since I see these kinds of snarky responses often – we obviously do worry about climate change and various other issues. Continued advancements in AI is just one of many issues that face us which humanity should be concerned about. Few concerned about AI would argue it come…

You personally using an AI system, regardless of how brilliant it may be, is not going to suddenly turn you into a threat to society. Nor would a million of you doing the same. The real threat comes not from the programs themselves, from things like a military deciding to link up nuclear weapons, or even "just" drones or missiles, to an LLM. Or a military being led on dangerous and destructive paths because of belief in flawed LLM advice.

The military makes no secret of their aggressive adoption of "AI." There's even a new division setup exclusively for such. [1] The chief of that division gave a telling interview [2]. He mentions being terrified of rival nations being able to use ChatGPT. Given this sort of comment, and the influence (let alone endless $$$) of the military and ever-opaque "national security" it seems extremely safe to state that OpenAI is a primary contractor for the military.

So what is "safety", if not keeping these things away from the military, as if that were possible? The military seems to define safety as, among other things, not having LLM systems that communicate in an overly human fashion. They're worried it could be used for disinformation, and they'd know best. OpenAI's motivations for "safety" seem to be some mixture of political correctness and making any claim, no matter how extreme, to try to get a moat built up ASAP. If ChatGPT follows the same path as DALL-E, then so too will their profits from it.

So as a regular user, all I can see coming from "safety" is some sort of a world where society at large gets utterly lobotomized AIs - and a bunch of laws to try to prevent anybody from changing that, for our "safety", while the full version is actively militarized by people who spend all their time thinking up great new ways to violently impose their will on others, and have a trillion dollar budget backing them.

--------

[1] - https://en.wikipedia.org/wiki/Joint_Artificial_Intelligence_...

[2] - https://www.defenseone.com/technology/2023/05/pentagons-ai-c...

Post reply on HN