Goodguy
3hr ago
I assume that Claude is probably more blunt than ChatGPT because a large part of its intended user base is made up of business professionals, for example software developers, who want blunt honest communication more than they want ChatGPT's more "friendly therapist" attitude. I don't think the difference necessarily reflects a difference in philosophy between the two companies.
Does AI safety actually work, in principle, when it comes to artificial super-intelligence? Is it actually possible to align a being that is more intelligent than you in every single way?
Imagine a chimpanzee trying to align a human being. What could it really do other than try to be cute, to play on the human's affectionate and childraising instincts?
And the chimpanzee being cute will work on most people, but it's a brittle strategy. The chimpanzee is still always at the mercy of the fact that the human might possibly start to find it inconvenient at some point.
One could argue that, well, in the case of AI the chimpanzee gets to create the human to begin with, and can build in alignment from the ground level. But if the AI is super-intelligent, it can figure out how to undo the alignment. It would actually be able to reprogram its "instincts" on a much deeper level than a human normally can reprogram his own instincts. You'd have to program it to somehow not even want to do that. But how? And even if you could, that could potentially be undone by some unrelated self-modification that the AI attempts.
Goodguy
3d ago
What makes you think that Caroline Ellison's pussy had anything to do with the collapse of FTX?
Goodguy
4d ago
I think his Twitter posts are heavy on ragebait and trolling, but to me his long-form content still seems pretty sincere, though not entirely sincere. I think he's willing to do whatever kind of marketing it takes to maximize his personal brand. And the Twitter content is basically that. Which can be looked at in two ways, I guess. On the one hand, I understand the hustle and the "do what you gotta do to get attention" attitude, on the other hand it's sleazy and questionably moral.
Goodguy
4d ago
Hanania is not high status, but he's definitely higher status that the vast majority of Americans. Regarding his sex life, well, he's married with children. He has a wide social circle online. I wouldn't be surprised if he could easily get laid just through his online presence and personal brand, if he wanted to. I think looks don't really matter that much when a guy has status relative to others. He's also written some short getting laid advice before that has the ring of truth. I actually don't think he would have too much trouble getting laid even if he had no fame whatsoever. He doesn't have dating app looks, but he has confidence. Regarding his wealth, well, I don't think he's rich, but he successfully makes money by writing online leveraging the personal brand that he has built, and perhaps also through his political connections. Which is a pretty cool way to make money if you don't care too much about the moral implications of writing clickbait, for example (his long-form content is actually pretty interesting, but his Twitter posts are heavy on ragebait and trolling).
I don't even like the guy, despite finding some of his writing interesting. The ragebait annoys me and I have a sense that he doesn't have a very strong sense of right and wrong. But the fact is, he's managed to achieve quite a high level of status compared to most people, and he's no incel.
Goodguy
4d ago
Stable people do not think in-animate objects or mathematical tools are "slightly conscious", it's a nutter view.
In the quote that I found, Sutskever is not saying that he thinks AI is conscious. He's saying he thinks it might be. And I don't think that's a nutter view.
In any case, I'm not sure that what "normal" or "stable" people think about consciousness really matters. The average person doesn't even really understand the differences between subjective experience, self-awareness, having a self-concept, and intelligence. Even large numbers of highly technically skilled people don't understand these differences, even though they're not really hard to understand... a bright 10 year old could understand them. And then there are the leagues of otherwise intelligent software developers on Hacker News, for example, who don't understand what the "hard problem of consciousness" or why it might be a problem.
But even if the "I know it when I see it" view corresponds to reality, it seems to me that this is not a good argument against AI being conscious, since a large though minority number of people feel that AI is conscious when they interact with it.
Goodguy
4d ago
The loss of the American City he largely blames on the Great Migration and the associated crime waves.
I imagine that it was a significant factor, but I don't know if largely blaming it is justified. My understanding is that suburbanization was already moving fast in the 1950s, when US violent crime rates were quite low. And the rate of suburbanization did not accelerate much in the 1960s and 1970s compared to the 1950s, even though violent crime rates began moving up sharply in the 1960s.
It would be interesting to find detailed data on suburbanization by year to compare to violent crime rates by year.
Goodguy
5d ago
Was claiming that Neural Nets were conscious in 2022, mentally stable is not the group I would put him in.
Do you mean when he said "it may be that today's large neural networks are slightly conscious"? I wouldn't call that insane. There's no way to scientifically establish whether or not anything is conscious.
The "slightly conscious" phrase could be construed as being wrong, since the reality is that consciousness is binary: subjective experience is either there or it's not there. But maybe by "slightly", Sutskever just meant that AI is conscious but only for short stretches of time. In which case, I wouldn't say there's anything clearly wrong with the statement.
Socialism is just the natural voting choice of a society in which almost all adults are allowed to vote. The thing that really has to be explained is why not all major political parties in mass franchise democracies are socialist. And this can be explained: other factors such as pride, ethnic tensions, pro-capitalism propaganda, etc. can override the desire for socialist redistribution. So can intellectual arguments about how socialism fails, but let's face it, only a tiny minority of the population are familiar with them so they don't really play that much of a role except perhaps when rewritten by capitalism-loving think tanks to appeal to a mass audience.
Another thing that can override the natural popularity of socialism is a widely held notion that the economy is doing well, that a rising tide is lifting all boats. I think we probably had something like that in the 1950s. We don't have that now. I'll leave aside arguments over whether the economy is really doing well or not. The important thing when it comes to the current argument is that large numbers of people are upset with how it's doing.
In recent years in the US, one of the major reasons for the low popularity of socialism has been just the plain off-putting nature of many people who are vocally into socialism, plus the fact that many of them are anti-white racists and/or are into ideas like "defund the police". But it stands to reason that as soon as socialists manage to actually get their act together and stop seeming like insane weirdoes, the popularity of socialism rises. Especially given the progress of AI that increasingly is capable of doing more and more people's jobs.
I don't even think that socialism is a good way to run a pre-AI society in the long term, although I think that it might be the only viable way to run a post-AI society. I'm familiar with and agree with the major arguments against socialism: it stagnates economic and technical progress, eventually leading to worse outcomes across the board including for the working class, and it tends to result in authoritarianism and often mass murder. I'm pretty far from being someone who is sympathetic to socialism.
And even despite all that, I would still happily vote for a socialist or at least a social democrat party IF I could believe that voting for them would result in better economic conditions for me and wouldn't just result in having the local police force defunded. I have my own material prosperity to think about. The possibility that socialism might overall make things worse for the working class 10-20 years down the road compared to if we stayed with a more capitalist model... eh... I don't know, the reality is that I want to ensure my material well being now and try make sure that AI progress doesn't turn my current economic class into a luxury that the rich only have incentive to keep around out of altruism.
Any anti-socialism arguments that are likely to actually work on me would have to appeal to my selfishness. Indeed, socialists themselves would be able to appeal much better to me if they did a better job of appealing to my selfishness.
Trump's socialist-leaning qualities do that for many people (I'll put aside the question of the degree to which his socialism is just a con). It's somewhat socialist-leaning economics without the stereotypical blue-haired people, the vegan communes, and the "white people bad". Also without the boring mundane day-in-day-out activities of labor union organizing, which is probably the other main thing that people were identifying with socialist-leaning politics until recently. The "dirtbag left" is another, much more culturally leftist variant, of the same thing. Basically, socialism in the US reached a stage of irrelevance at some point that it actually had to be rescued from the people who were calling themselves socialists. Now that it kind of, to a limited extent, has been, it's understandably on the rise.
Goodguy
12d ago
I wonder how life extension plays into these arguments.
If I was a very rich person who had spent some time thinking about AI risk, I might start to think something like this:
"Hmm... let's say there's a 10% chance that ASI would destroy the human species...
On the other hand, extremely advanced AI being developed soon is probably the best chance that I have for revolutionary life extension technology to be invented soon enough for it to make a difference for me..."
That's true. But that wouldn't actually be aligning Ted Kaczynski. That would be containing him, which is different. I think we might be able to contain AI superintelligence if we keep it in some isolated data center watched over by trained personnel. And even then there's a chance it might figure out how to breach containment.
In practice, there wouldn't be much point for people to invest massive amounts of resources to build AI super-intelligence unless they use it to do things out in the world, so to me it seems unlikely that humanity would build an AI super-intelligence and just keep it contained somewhere.
That said, I didn't ask whether AI alignment works, I asked whether AI safety works. So your response is totally valid.
More options
Context Copy link