@VelveteenAmbush's banner p

VelveteenAmbush

Prime Intellect did nothing wrong

3 followers   follows 1 user  
joined 2022 September 05 02:49:35 UTC

				

User ID: 411

VelveteenAmbush

Prime Intellect did nothing wrong

3 followers   follows 1 user   joined 2022 September 05 02:49:35 UTC

					

No bio...


					

User ID: 411

Hi everyone, so happy this site is up and running and hope we succeed in bringing over the community!

Can we have a "new site questions and feature requests" thread? I come bearing content along those lines.

Counterpoint: No, we're a mostly aligned blob of people who agree on most issues, and the impression that we're headed for a civil war is an artifact of being extremely online, where insane people of all stripes are disproportionately present and energetic.

Make a falsifiable prediction with a date attached to it.

I am hoping a computer person can extract the exact CSS or whatever from an old.reddit.com comments view and implement the exact same thing here. The tiny text, oceans of whitespace and unfamiliar UI widget placement here makes it difficult for me to use this forum.

I'll even be nice and say we don't have to count the BLM riots, as if we did my prediction would already be realized.

This is where you give it away that you're just predicting that our status quo (of periodic political events accompanied by politically motivated crime) will continue. You said "There cannot be one nation ruled by two tribes. We will divorce or we will come to blows, eventually. One will become two or two will become one." You think a couple more political riots in the next decade is all that it takes to substantiate this apocalyptic vision? Come on, either tone down the rhetoric or make a prediction that justifies it.

If one accepts the vegan frame that animal lives have moral value, then one argument is that livestock wouldn't exist if we didn't farm them, and while that makes us omnivores responsible for their deaths, it also makes us responsible for their births. Are their lives really so bad that they'd have been better off never having been born? For some types of factory farming, this is probably true, but not for all. A world of vegans is a world where cows would probably be extinct, or close to it.

Would it be possible to extract the exact CSS or whatever from an old.reddit.com comments view and implement the exact same thing here, ideally as the default? The tiny text, oceans of whitespace and unfamiliar UI widget placement here makes it difficult for me to use this forum.

This is pretty clearly a woman.

By what definition of woman?

I think the whole trans discourse of whether someone "is" a woman is fairly hopeless. The true request is for admission to the social institution of womanhood. And I think a lot of people who are willing to treat the person as a woman -- use her preferred pronouns, not object if she uses the women's bathroom, not give her shit for wearing a dress and makeup no matter how incongruous it seems in proportion to her profile or vocal intonation, etc. -- nevertheless balk at agreeing that she is a woman.

Accommodation is a natural and understandable request. We're used to it in a variety of contexts. It has been a common form of social compromise for longer than civilization has existed. We're used to acting as though there is nothing different about people in wheelchairs, people with congenital deformities, people with dwarfism, people who are extremely old, people who obviously have a terrible disease, etc.

But if someone in a wheelchair wanted me to say they could walk... if someone with dwarfism wanted me to say they were six feet tall, if someone in her late eighties wanted me to refer to her as a 25-year-old... that would be hard for me.

Unfortunately, agreeing that someone is a woman is an unavoidable part of granting them access to the institution of womanhood.

So, it's a pickle.

"Gender role" suffices for that purpose and isn't contentious.

So their manager asks them to do something about bias, and they apply the laziest possible hack.

I actually have a different impression: most of these professional ML researchers and engineers genuinely wish they could serve up a model that provides politically correct responses, because politically correct responses are also commercially correct, and everyone wants to make money. Probably the main reason a bunch of giant and amazing Google models aren't made available to the public via API is because of the risk that they might say or display something politically incorrect, and certainly some fraction of the user base (especially tech journalists lusting after those sweet engagement metrics) will try to bait it into doing so.

So there's ample incentive to solve this problem "the right way," and the fact that so far all we see are cheap hacks and opacity is because no one knows how to solve it the right way, or even if it is solvable the right way at all, even in principle, with the technology we have today.

Part of the problem is exactly what makes these models so exciting to begin with. They can notice things, they can extrapolate from training data, they can make analogies and they can roll with out-of-sample prompts, and they develop all of these amazing abilities ex nihilo, from a largely uncontrollable black box made of inscrutable matrices gently nudged in the direction of data.

The other part of the problem is that political correctness isn't a well defined or static problem. It is a messy social problem, involving subtle adversarial factional games, sort of like fashion.

And these two halves of the problem compound with one another. It isn't enough to generate a black person one time in X -- you have to define X, you have to solve this equation for all possible identities, and you have to then translate this equation into every conceivable fact pattern that the user will (adversarially) use to challenge the model. If you want to generate a picture or story involving a policeman arresting a criminal, it is fraught whether you make the policeman white or black, whether you put him in a wheelchair or not. Should the model generate trans women? If they're visibly identifiable as trans women, are you making a minstrel caricature to further the stereotype that trans women look like men in dresses? If they aren't visibly identifiable, how is one to know they are trans at all, and that you haven't committed the deadly sin of erasure? Should black women look like white people but with a darker skin tone (and draw criticism for e.g. straightening her hair, itself a political minefield), or should you make them look recognizably phenotypically black in terms of facial features and hair (and draw criticism for reinforcing a stereotype)? If both murderers and NBA stars are disproportionately likely to be black, does the model need to recognize that murderers are bad and NBA stars are good and apply its distortion of the underlying distribution only to the bad category, i.e. return mostly white guys for criminals but mostly black guys for NBA stars? How is it to know? And when ideological opponents start to stress-test these categories and ask for a thuggish NBA player or a corrupt President, should it reverse the categories? What about middle grounds, like an "aggressive" NBA player, or a "desperate, nonviolent" criminal? We even have minor culture wars about the perceived race of robots.

and gas/electricity prices are actually coming down at the moment due to windfall taxes being imposed on the producers and changing the mechanism through which electricity prices are set with any monies raised being used to subsidise households.

This seems to be a case of mistaking the map (price) for the territory (supply). Distorting prices downward via government edict is going to result in overconsumption and increase the risk of shortages.

I know almost nothing about energy forecasting, but naively could the lower futures prices reflect an expectation of rationing or price ceilings as well as they could reflect an expectation of abundance?

That admission is going to have to happen one way or another, if the facts demand it.

Yes, I think this sort of "prosaic alignment" solution is likely to solve all of our consternations about AI capabilities at least as well as a human intelligence could... in the long term. Eventually, you wouldn't even have to talk about portraying X ethnicity negatively, you could just say "and make it politically correct" and the AI would understand those rules better than any individual. For the time being, though, Dall-E has a hard enough time drawing a complex but coherent picture, much less enforcing its conformity with protean standards of political correctness.

Worth pointing out that OpenAI tried this sort of "prosaic alignment" approach to its so-called diversity filter. It appears to append stuff like "black male" or "hispanic female" to some proportion of prompts that it believes call for the depiction of a person. It has been vigorously panned by the community, because it has unintended negative effects on many prompts. Gwern did some sleuthing on why a complicated prompt for a picture of a cowboy at a certain angle in certain lighting etc. returned a bizarre misfire, and eventually discovered that the same prompt with "cowgirl" instead of cowboy worked flawlessly -- seemingly implicating the diversity filter in the original prompt's total failure.

Hilariously, OpenAI's approach here was discovered by asking for stuff like "A person holding a sign that says" -- and then you'd often get a picture of a sign that says "FEMALE" or "BLACK". So there's a degree to which adversarial prompt construction can overcome attempts at coercive prosaic alignment, at least using current techniques.

It also doesn't know our delicate rules about when it's socially appropriate to re-gender or trans-racialize the subject of a picture. It's weird if prompts for Princess Zelda return a black or Asian Zelda, or if prompts for George Washington return a colonial-era woman in a white wig. Maybe we'll accept that sort of thing by the time Season 10 of Bridgerton comes out, but I don't think we're there now, and it would take a pretty advanced AI to figure that stuff out.

but if one starts from the assumption that disparities in outcomes reflect the magnitude of discrimination

One can always prove too much if one starts from false assumptions.

But, we still have to deal with the fact that we live on a planet with a finite amount of space and a finite number of physical resources, so it's hard to see how we could ever get to true post-scarcity.

Why would we be limited to this planet?

No way. The environmentalists can't stop fracking, and even nuclear power is having a renaissance. We aren't going to accept a dismal Malthusian future like OP suggests out of a desire to leave all of the rest of the mass-energy of the universe pristine.

If Germany took orders from the US, they wouldn't have built the Nord Stream 2 pipeline in the first place.

Why should the US expect much help from Europe on the main front against China in the Pacific?

Who's expecting that? Europe's NATO expenditures speak for themselves, and even those are defensive in nature. Any European participation in a Taiwan defense will be to support a branding of multilateralism, not for their direct material support.

I'm American and I can't make any sense of our foreign policy strategy in regards to Russia. It all seems like dick-waving with potential nuclear consequences. What do we even get if we "win?"

It doesn't seem so hard to understand to me. Russia is openly defiant of the Western order and props up our military adversaries like North Korea, Syria and Iran. Even Turkey, ostensibly a NATO member, became a patron of Russian defense systems. Adding insult to injury, Western Europe voluntarily made themselves dependent on Russian energy and starved their own militaries of funding and capabilities necessary to pull their weight in furthering the US's foreign policy goals.

Now, Russia is isolated, its military has been revealed as an underperforming anachronism, its brand has suffered in the arms marketplace, the Russian energy link to Western Europe has been perhaps permanently severed, Europe is committing to nuclear energy and US LNG imports as fast as it can, Putin is facing domestic political troubles, NATO is expanding to include wealthy Nordic nations that have resisted membership for decades, China has distanced itself from Russia and is probably rethinking its ability to take Taiwan by force, Russia's ability to maintain its support of its military client states (e.g. Azerbaijan) is faltering, and all of this has been achieved with no more cost to the US than a few hundred billion dollars of military equipment and some intensive military consulting with Ukraine. Europeans will shiver through the winter and Ukraine's streets are red with blood, but those are other people's (and peoples') problems. Even if Putin uses nuclear weapons, it seems unlikely that he'd target US territory, it's an open question how effective they would be, and he'd open the door for a much firmer response that would further accelerate all of the above.

If you're an unsentimental partisan of US interests, what's not to like?

people may have a very quixotic idea of their own or others' bono

By far my favorite phrase of the day.

one beautiful thing of the hypothetical true masterminds would have been to make the ultimate guilty party the radical wing of German Greens or some other environmental nuts.

But these groups are surely even more ineffectual than the Mexican cartels, no? It takes a lot less sophistication to bomb an undersea pipeline than to put a man on the moon or whatever, but surely a lot more than it takes to block traffic, and I see the latter as more within the capabilities of enviro-nuts.

Edit: Seeing elsewhere in the thread that the pipe was only 350 feet underwater. Maybe this operation would have been within the capabilities of the enviro-nuts after all, and could have been disguised as such -- but the size of the actual explosion seems like it would have been beyond them.

How fast can they scale up LNG shipments from the US and get serious about domestic fracking? Surely not fast enough to save them this winter, but I would (perhaps naively) think this is no more than a two to three year project before they start seeing some of the benefits.

Residential heat pumps are probably the best short-term answer.

Jim Geraghty is not causally involved, obviously. His role here is as an analyst and commentator at best. Still, his analysis makes sense to me.

Their punishment? Nothing.

They seem to be punishing themselves rather effectively. Kind of incredible, really, that they're still toiling away at zero covid.

I think the US foreign policy establishment is actually doing a pretty good job of threading the needle and avoiding the risk of nuclear escalation. It is possible that Putin will resort to a tactical nuke, but (1) I suspect it'll be less effective than the conventional wisdom has it, and may actually deflate some of the mystique around nuclear weapons and lower the odds that they're used in the future, (2) I'm sure we'll have a very sharp response but will avoid retaliating with nuclear weapons of our own, and if NATO is not directly kinetic on (actual) Russian territory and Ukraine constrains its kinetics only to military resources on Russian territory, then Russia will have every incentive not to use nukes outside of Ukraine, and (3) breaking the nuclear taboo will make it even harder for Russia to come back from its isolation even after this has all blown over.

Sure, a fireball in Ukraine would suck, and the mountains of rubble and shattered bones in Ukraine already sucks, but neither really damages US interests.