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

The good delusion: has effective altruism broken bad?

economist.com

Post reply on HN
Print view View original

581–590 of 619 posts

Re: The good delusion: has effective altruism broken bad?

#581

Post by Viliam1234 »

thegeomaster wrote: post #279
tptacek wrote: post #137

My "Effective Altruism Is Not A Cult" t-shirt is prompting a lot of questions already answered by my shirt. I've dug into this stuff over the last couple days, because I had previously understood Effective Altruism to be something akin to like a pledge drive for Charity Navigator's best-reviewed charities or something, and it turns out it's nothing of the sort. In fact, if you read some of the commentary from EA insi…

>maybe EA is fundamentally Rationalist? The EA forum in fact uses the forum software developed by LessWrong, in very close collaboration with them (it's not open source or anything). In general it looks like the two organizations have a lot in common, and LessWrong is a poster child for the rationality movement.

> it's not open source or anything

GPL3 actually https://github.com/ForumMagnum/ForumMagnum

> In general it looks like the two organizations have a lot in common

Yes, see https://en.wikipedia.org/wiki/Effective_altruism#History

Re: The good delusion: has effective altruism broken bad?

#582

Post by danaris »

naasking wrote: post #556
danaris wrote: post #546

Earlier quoted context omitted.

You don't need to have a deep and complex understanding of human sentience to be able to put some lower bounds on what it means to be sentient. I don't know about sentience , as such; however, I believe there are at least two bare-bones prerequisites to an AI being conscious as we understand the term: 1. Continuity of input: The AI needs to be constantly "on", constantly receiving input of some sort, and constantly a…

> Continuity of input: The AI needs to be constantly "on", constantly receiving input of some sort, and constantly able to produce output, rather than being strictly limited to producing discrete responses to discrete stimuli. Why can't it be conscious merely while the computation is running? You aren't constantly on, you're ostensibly only conscious 16-20 hours every day, while you're not sleeping. > Continuity of l…

I think that both of your arguments against my points reflect a view of "consciousness", as we humans experience it, that is much more binary than I believe it is. (Either that, or something of a misunderstanding of my points.)

Even when we sleep, our brains are still active. They are still processing input and consolidating information.

Even when we meditate, our brains are still active. I don't know why you think Zen Buddhists being conscious during meditations in any way refutes my statements; I think it supports them.

A system like GPT-3 never receives continuous input. It receives discrete input: one prompt, then another; a static dump of training data.

To the best of my knowledge, none of the extant major ML systems do continuous learning. They are given training data when they are first set up, and they may be given further rounds of training later, but they have no ability to actively, continuously receive input and consolidate data.

We do not have the ability to genuinely shut off our brains. Not in the way that we shut off computer programs. And I believe that that continuity is a fundamental prerequisite to something being "conscious" in any way that we would recognize. Furthermore (to loop back to the original topic of "the dangers of AGI"), I have absolutely no doubt that any AI that we could create that lacks continuity would never have the ability to represent a serious threat to us, because all we'd have to do to be safe from it is to not activate it again as soon as we detect trouble.

Re: The good delusion: has effective altruism broken bad?

#583

Post by sooheon »

rsj_hn wrote: post #234
sooheon wrote: post #229

Earlier quoted context omitted.

> "My altruism is more effective than anybody else's, so I'm going to steal everybody else's money so that I can put it to better use." > How do you stop this shit? If you ask yourself why you think "this shit" is bad, at root there will be a consequentialist reason. Deontological ethics is consequentialism by heuristics.

But if you ever try to put any consequentialist theories into practice, at root there will just be a set of deontological heuristics guiding your behavior. Consequentialism is just deontology in drag.

You're right. We are unable to be full consequentialist because the computational cost of seeing all ends is infinite, and unable to be full deontologist because competing heuristics are arbitrarily instantiated and mutually contradictory. Even if we're stuck in the middle forever, I know which way I'd strive.

Re: The good delusion: has effective altruism broken bad?

#584

Post by cactusplant7374 »

coldtea wrote: post #308
cactusplant7374 wrote: post #294

Earlier quoted context omitted.

The question is why do programmers believe they know more than Chomsky? The AI we have now is engineering and not science. Don’t you need science to create AI? Don’t you need to understand something fundamental about the brain?

I don't believe that much in any realistic timeframe from AGI, but I don't think you "need to understand something fundamental" to create it. You can create almost anything empirically, without knowing the science behind it at any fundamental level. Evolution, for example, didn't understand anything (as it's just a process). It wasn't someone with an understanding of brains at the scientific level that built our brai…

If I understand you correctly, your sentiment is that humans can stumble upon and solve the most complex problem on the planet?

OpenAI claims to be working on AGI. Nothing they have built is even tangentially related.

Re: The good delusion: has effective altruism broken bad?

#585

Post by cactusplant7374 »

olalonde wrote: post #399
cactusplant7374 wrote: post #294

Earlier quoted context omitted.

The question is why do programmers believe they know more than Chomsky? The AI we have now is engineering and not science. Don’t you need science to create AI? Don’t you need to understand something fundamental about the brain?

Not necessarily. Volta invented the battery almost a 100 years before the electron was discovered. Also, planes fly very differently from birds.

Those things are a billion times less complex than human cognition.

It’s also interesting your remark on birds and planes. Chomsky uses a similar analogy to say they are dissimilar: can a submarine swim? Well, depends on what you mean by swim.

Re: The good delusion: has effective altruism broken bad?

#586

Post by onetimeusename »

naasking wrote: post #533
onetimeusename wrote: post #29

Earlier quoted context omitted.

that's a good way of putting it. I kind of ignored a lot of what they said but again, just like the rationalists who pretty much were all of the EA people, they came with biases they refused to admit they had. like for example, if a family member of mine needed money for a cancer treatment that had a 50% chance of working versus giving that money to save the life of someone I don't know on the other side of the plane…

> like for example, if a family member of mine needed money for a cancer treatment that had a 50% chance of working versus giving that money to save the life of someone I don't know on the other side of the planet, I will consistently try to save my family member. That is how most people would operate. I just want to point out that: 1. Just because this is how most people operate does not mean this is the ethical cho…

yes I am biased if bias means not weighting all people equally on moral grounds. I am not sure why I should. If someone did that I think it's interesting what conclusions they might draw about how to spend their money or what job they might choose. It destroys much of the point of having a family. You would have to weight random strangers with equal status to your own children. So the needs of random strangers would almost always outweigh the needs of your children assuming they are healthy and you live comfortably. It would not surprise me if EA people thought families were immoral though.

I don't think assuming an equal weighting is being unbiased though. It could be a result of a political bias. As I said I don't know why people should assume this. The digital era of social media and digital banking enable this viewpoint so I think that is another angle to discuss regarding how to weigh these issues.

But the specific bias I was accusing some in the EA community of having is virtue signalling as having superior ethics, morals, and intelligence. I think that is a real bias. I don't want to accuse everyone in the community of not being altruistic but claiming to be an effective altruist could certainly be a career enhancer if you are an ethicist. Also, my experience with this community is some of them come off as misanthropes. A misanthrope might be motivated to argue elaborate contrarian viewpoints in order to cause harm or annoy people.

Re: The good delusion: has effective altruism broken bad?

#587

Post by rsj_hn »

sooheon wrote: post #583
rsj_hn wrote: post #234

Earlier quoted context omitted.

But if you ever try to put any consequentialist theories into practice, at root there will just be a set of deontological heuristics guiding your behavior. Consequentialism is just deontology in drag.

You're right. We are unable to be full consequentialist because the computational cost of seeing all ends is infinite, and unable to be full deontologist because competing heuristics are arbitrarily instantiated and mutually contradictory. Even if we're stuck in the middle forever, I know which way I'd strive.

Real heuristics, of course, are not arbitrary, they are the result of trial and error processes where societies and people with differing moral systems compete with each other, proving the fitness of their heuristic by surviving while less fit heuristics fail.

I'll go with the results of trial and error over long time horizons rather than blank slate theoretical calculations any day.

Re: The good delusion: has effective altruism broken bad?

#588

Post by FeepingCreature »

michaelmrose wrote: post #549
FeepingCreature wrote: post #328

Earlier quoted context omitted.

Maybe we should try to acquire the tools before we build systems much smarter than we are

Current AI seems to be just a box of math churning out correlations that we can put to effective use but not understand wherein the ability to actually act is entirely reliant on us deliberately wiring the machine into the world in a way it can even have negative effects or any effect at all wherein said negative effects are entirely and trivially predictable and avoidable. This is true whether its a robot that accid…

None of this seems to argue against focusing on safety tools now, given the enormous downside risk. And maybe you're wrong about the probable decades.

The possible outcomes here are "probably we waste a lot of money" and "the world ends."

Re: The good delusion: has effective altruism broken bad?

#589

Post by FeepingCreature »

pvg wrote: post #510
FeepingCreature wrote: post #326

Earlier quoted context omitted.

The atom bomb was once a hypothetical construct. The Einstein-Szilard letter was talking about a danger that they had no way of proving was even possible. Proof is irrelevant outside of mathematics; the only relevant question is cost to expected benefit - or expected harm. With dangerous research, when a conclusive demonstration of possibility is achieved, it is usually too late to do much in the way of prevention. T…

The atom bomb was once a hypothetical construct. The Einstein-Szilard letter was talking about a danger that they had no way of proving was even possible. This isn't an accurate representation, really. The letter was written exactly because it was clear that a fission bomb is very likely within the short-term reach of current technological and industrial capability. This is very much not the case with AGI.

That's a difference of degrees, not kind. You can always ask for a higher class of proof.

Re: The good delusion: has effective altruism broken bad?

#590

Post by wolverine876 »

UncleMeat wrote: post #447
PebblesRox wrote: post #434

Earlier quoted context omitted.

"Maybe it's some kind of horrible drug interaction with Rationalism, or maybe EA is fundamentally Rationalist?" I have always associated EA with the rationalist movement, since I came across it via reading rationalist stuff. If you define EA as "figuring out how to deploy a given set of resources so as to do the most good," your conclusions are going to depend on your value system: how do you define and measure good?…

One of the problems I observe in forum posts from these communities is that it is fun and cool to think differently than others. "What if fighting poverty now is actually bad and preventing the AI apocalypse later is actually better?" Then you get to feel superior to everybody, even those goody-two-shoes people down the street that donate a lot to the food bank. You know the secrets. It is the same kind of thinking t…

Like much tech wealth wisdom, it's rationalization of their megalomania, their self-serving desire for power.

Humans naturally desire power to varying degrees, but we have other desires and we have other things that limit us (e.g., our consciences, or Adam Smith's 'impartial spectator'). With enough real power, you can ignore these things and others will not only fail to check you, but will encourage you. Power corrupts.

Post reply on HN