aqouta
Friends:
User ID: 75
aqouta
1d ago
I'm not sure how I fall on this topic but to steelman the pursuit from a utilitarian lens the suffering of a few AIs to show that AIs do feel pain could act as strong evidence that we shouldn't torture many many more AIs. If this prevents one of the labs from stumbling accidentally on a combination of weights they don't believe to be meaningful that they can push when the model misbehaves to make it behave more and comercializes it without knowing they just added a torture feature to their product then perhaps billions of times more harm will be done than if this can convince them not to. Now of course there's the chance that some lab will anounce "we have at least added the automatic torture feature to our product from the great paper "do no add the automatic torture feature".
aqouta
8d ago
He'd start by explaining why you need to do some basic physics experiments and measurements which you'd fuck up and the communication would be annoying, you'd have to keep hitting "allow" so that he could talk to the assistant. Eventually you'd just let him talk whenever he wants and soon it would be clear that he needs to be on site for the experiments. Just observe anyone using these tools man, we're not keeping the ai in the box.
aqouta
9d ago
Politically neutral outfit for a rapper.
Yeah, having known about this before hand and not seeing it in the OP felt pretty sus.
aqouta
9d ago
this is a fair point but I'd counter that even without the AI use case there are business incentives to create and curate well tagged data for marketing and social media uses. I don't want to argue too much in defense of specifically how late the transformer model would need to be discovered to make FOOM possible, the path didn't go as Yud predicted. In theory this there could be some other architecture that would have different features but that's unfalsifiable.
aqouta
9d ago
I can't really argue against your pessimism points. I'm mostly just hoping I'm either wrong about something fundamental or scaling stops working sometime very soon for reasons I don't currently understand, which you mention. Obviously, at least to us, that cope is seeming less realistic with every bit of news coming out of the labs. I do think that if we managed to pull off a pause and everyone just had access to current gen models with maybe some cost savings we could still use them to accelerate cancer and aging research relatively safely. We've barely even started putting the currently available models to use.
aqouta
9d ago
As far as being wrong about the fundamentals of how AI was going to be developed I think the most wrong thing he was confident about was the idea of FOOM and in general not predicting this multi-year phase we're currently in as training and scaling continue to produce steady returns in capabilities. Ironically if the tranformer architecture came out in like 2030 instead of 2017 and compute was abundant relative to today we might have seen a FOOM. Albeit I concede on a lot of the specific substrate stuff both he and Hansen talked about were well off. In their defense nearly no one else registered a prediction at all so while we can hold this against them I object to raising anyone else's predictive power relative to them.
His stuff on alignment in general all holds up pretty well and I'm not so sure you can really accuse him of believing rather than hoping that there are stops on the train. He seems to have correctly intuited that trying to add breaks was a monumental task from a remarkable distance.
aqouta
9d ago
Apparently slaughtering Persian schoolgirls for no reason is a-ok but aborting a 2 month fetus is not.
No one is in favor of slaughtering Persian school girls on the first order so it's just not virile memetics.
aqouta
9d ago
Well, actually when you put it like that they are still building the datacenters right? So they aren't even pulling back in capex which makes your narrative even more ridiculous. Thanks for the contribution.
aqouta
9d ago
If only they had long records of writing on the topic going through their beliefs and positions in detail that you refuse to recognize.
aqouta
9d ago
I think giving the kid any amount of fent would have the exact same effect as whatever they gave him.
aqouta
9d ago
And there isn't any possible evidence that could exist that could disabuse you of that impression because pausing unavoidably causes less capex spend. You're epistemically sealed against the straightforward explanation for why the people are doing the things obviously motivated by the durably stated beliefs over many years.
aqouta
9d ago
The embittering part of it all is that I wish I could agree with the accelerationists, it'd be so beautiful, and I'm not even full Yudd in that I think there's a very good chance Alignment ends up being tractable and even might go well automatically. But I just can't stomach ten or twenty percent chance that everyone dies. It can still be beautiful even with a pause, we can still cure cancer and fight back against aging and probably end all toil. But it'd be so nice to not have to feel trepidation about this.
aqouta
9d ago
I don't think this models the feminist side very well, these are the same people who will defend a woman's murderer and hope he doesn't get locked up forever. I think the simple explanation is that to many, and women are more susceptible than men, empathy fuses with sympathy. They experience putting themselves in the shoes of the person being spoken about, the murderer, and they immediately feel the walls closing in. It's not real empathy, they aren't imagining that they're the kind of person that would kill their kid they're imagining that they are themselves in a situation where they have just killed their kids, which they know would have been impossible for them to do in their right mind and thus it would be unjust to expect them to take responsibility for the actions. Yes, this isn't workable reasoning and generalizes in such a way that everyone would be forgiven for every crime, but they aren't doing reasoning, they're doing sympathy and you can observe them doing it every time.
aqouta
9d ago
The safetyists have worked hard to not be associated with the previous wave of "AI bias" people and the more recent wave of datacenter luddites. There were discussions on whether they should try to harness the popular, but retarded, backlash but wisely decided that polarizing the concept of Safety work was far too important for some cheap support.
aqouta
9d ago
There is very obviously not a single thing that could ever satisfy you people. You're twisting literally calling their own sole product a risk to kill all humans in a marketing gimmick to secure regulatory capture. If they carved "No really we're serious about the risks" into their chests and then hung themselves you'd claim they were expecting to receive compensation in the afterlife.
aqouta
9d ago
I'm just glad that we're finally past the totally clueless people insisting AI is not even useful for coding and are on to the next stage of grief we have to drag people through on the way to acceptance that big Yud has been right the entire time. The faux cynical takes that the labs at the absolute cutting edge of this are fudding their own product to the dismay of the sitting president in some insane 8 dimensional chess move because they want to be regulated by Bernie Sanders is so plainly retarded that it can't possibly last very long.
aqouta
9d ago
In practice, AI safety is Anthropic, it's EA, it's Twitter activisits and 'explain how you will prevent your AI birdspotting app from discriminating against racial minorities' checklists and 'data center are using all our water' and 'Banning Home GPUs Is Not Dystopian' and putting Chinese-controlled bombs in datacenters. MIRI is irrelevant and I doubt trump has an opinion on them either way.
This is just straightforward outgroup homogenity bias. The AI safety people despise those other groups and intentionally worked hard never to associate with them.
aqouta
9d ago
To my mind, Anthropic is to AI Safety what the Soviet Union was to Communism: the obvious result of their doctrines when actually applied. It was founded to do AI Safely, by AI Safety extremists, and according to candidate reports it imposes incredibly constraining cultural interviews to make sure that nobody who isn't an AI Safety extremist can get a job there. They specifically filter out people who are visibly interested in improving AI capabilities. They have to offer services to customers to prevent more responsive and less AI-safety-dominated companies taking over their position, but they've been clear they only offer services to the extent they have to in order to do AI safety.
This is the case for them being the good guys, yes. Although I don't think they're as good as you say.
aqouta
13d ago
In context of above "conspiracies", maybe you are aware that for instance former Young Turks progressive Anna Kasparian is highly supportive of Owens theories. She endorses Owens, Carlson and is fully behind them. So by that metric I can say that It is no secret that DNC under progressives really like conspiracies
Anna Kasparian is of the opinion that the Democrats are effectively the same as the republicans and does not support them. Her an her cohost are quite insistent on this.
aqouta
14d ago
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
Socialism is founded on the lie of capitalist exploiting worker's surplus value and proposes in many ways to prevent that through worker owned companies or by taking the wealth of the capitalists and returning it to the workers. It's incoherent to talk about socialism in a post labor world. What you're describing is a system of redistribution.
aqouta
15d ago
The minimal moat they give you is the table stakes. There's no benefit to raising the bar for entry when everyone who was ever going to pay them has already ante'ed up. Like putting the condom on after sex.
How did this work out for the nuclear industry in the US? This is absurd, the federal government lately is not in a "give a couple operators a monopoly" state, it's in a "destroy any industry that makes the voters angry" state. The idea that this is a positive EV play is plainly just cynicism for the sake of cynicism. These are not companies whose major competitive advantage is app install base, they're companies whose major competitive advantage is billions of dollars in R&D in an industry where catchup mechanics are brutal with overseas competitors that are not at all subject to preferable regulations. There's just no actual way this works out to being a good EV play over pushing the frontier and hoping for a moonshot. Even if all the major players in all the major labs weren't on record for over a decade with these exact same concerns well before it was clear anyone was going to make any money in this industry this would be a completely retarded plan. "establish a regulatory regime where every company that could have plausible fronted the capital to compete has already done so" is not anything like a real cartel. What are they, afraid of netflix's entry into the sphere?
aqouta
20d ago
Apologies, when I critiqued socialists and you responded by saying class is real and hierarchical I had assumed you were taking the common socialist position that these hierarchies were unjust. You seem to agree somewhat with Marxists that class is real but you believe it is not unjust.
This is a coherent position but I'm not sure what you're trying to say about my observation that leftists want to squish one hierarchy, economics, and leave the other, social. Is it your position that the social and economic hierarchy are identical?
There totally are right wing bars. I've stumbled into them before even in Chicago. They're not so explicit, it's usually more local community first with the implicit assumption that the local community is right wing. But they definitely exist.
More options
Context Copy link