This weekly roundup thread is intended for all culture war posts. 'Culture war' is vaguely defined, but it basically means controversial issues that fall along set tribal lines. Arguments over culture war issues generate a lot of heat and little light, and few deeply entrenched people ever change their minds. This thread is for voicing opinions and analyzing the state of the discussion while trying to optimize for light over heat.
Optimistically, we think that engaging with people you disagree with is worth your time, and so is being nice! Pessimistically, there are many dynamics that can lead discussions on Culture War topics to become unproductive. There's a human tendency to divide along tribal lines, praising your ingroup and vilifying your outgroup - and if you think you find it easy to criticize your ingroup, then it may be that your outgroup is not who you think it is. Extremists with opposing positions can feed off each other, highlighting each other's worst points to justify their own angry rhetoric, which becomes in turn a new example of bad behavior for the other side to highlight.
We would like to avoid these negative dynamics. Accordingly, we ask that you do not use this thread for waging the Culture War. Examples of waging the Culture War:
-
Shaming.
-
Attempting to 'build consensus' or enforce ideological conformity.
-
Making sweeping generalizations to vilify a group you dislike.
-
Recruiting for a cause.
-
Posting links that could be summarized as 'Boo outgroup!' Basically, if your content is 'Can you believe what Those People did this week?' then you should either refrain from posting, or do some very patient work to contextualize and/or steel-man the relevant viewpoint.
In general, you should argue to understand, not to win. This thread is not territory to be claimed by one group or another; indeed, the aim is to have many different viewpoints represented here. Thus, we also ask that you follow some guidelines:
-
Speak plainly. Avoid sarcasm and mockery. When disagreeing with someone, state your objections explicitly.
-
Be as precise and charitable as you can. Don't paraphrase unflatteringly.
-
Don't imply that someone said something they did not say, even if you think it follows from what they said.
-
Write like everyone is reading and you want them to be included in the discussion.
On an ad hoc basis, the mods will try to compile a list of the best posts/comments from the previous week, posted in Quality Contribution threads and archived at /r/TheThread. You may nominate a comment for this list by clicking on 'report' at the bottom of the post and typing 'Actually a quality contribution' as the report reason.
Jump in the discussion.
No email address required.
Notes -
Finally, concrete plan how to save the world from paperclipping dropped, presented by world (in)famous Basilisk Man himself.
https://twitter.com/RokoMijic/status/1647772106560552962
TL;DR: GPU's over certain capability are treated like fissionable materials, unauthorized possession, distribution and use will be seen as terrorism and dealt with appropriately.
So, is it feasible? Could it work?
If by "government" Roko means US government (plus
vassalsallies) alone, it is not possible.If US can get China aboard, if if there is worldwide expert consensus that unrestricted propagation of computing power will kill everyone, it is absolutely feasible to shut down 99,99% of unauthorized computing all over the world.
Unlike drugs or guns, GPU's are not something you can make in your basement - they are really like enriched uranium or plutonium in the sense you need massive industrial plants to produce them.
Unlike enriched uranium and plutonium, GPU's were already manufactured in huge numbers, but combination of carrots (big piles of cash) and sticks (missile strikes/special forces raids on suspicious locations) will continue dwindling them down and no new ones will be coming.
AI research will of course continue (like work on chemical and biological weapons goes on), but only by trustworthy government actors in the deepest secrecy. You can trust NSA (and Chinese equivalent) AI.
The most persecuted people of the world, gamers, will be, as usual, hit the hardest.
To all people that think "the gubmint" can save us from AI I have only one question: what year it was when the US Government won the War on Drugs?
Note that drugs just make you feel better, they don't actually give you superpowers. And most of them kill you quite fast. Now imagine there was a drug that actually can make you 10x smarter, with no (immediate) adverse effects. Do you think it's something that could be suppressed, knowing what we know about the world around us? Do you think "just take the fascism to 11 and it solves everything" is going to help?
They always do. You see, that fascism wasn't rationally planned enough, now here's my proposal…
I think people who make decisions should have less credentials and more experience with different and unappealing facets of the world – and not isolated to enclaves of their distributed elite republic of conferences and respectable institutions. Regularized, so to speak. Ideally we should have the equivalent of Rumspringa for the aspiring PMCs. Imagine if every one of those do-gooders had to work for a year in a provincial Russian police office, then in a Haitian hospital, then something else along these lines. Perhaps the resulting selection and education wouldn't necessarily be very helpful, but it'll probably make things more interesting.
Seriously though I accept @2rafa's arguments, «compute governance» at least for new capacity is easy. I'm also more pessimistic because there's significant political will in favor of it, and crucially, no coherent and comparably fanatical opposition. We know how this went in Culture War battles.
When I was a kid and my father taught me how to play chess, the most useful lesson was this: "remember, in chess there are turns - after each of your turns, the opponent gets to do his. If you have a plan which only accounts for your side, it's going to survive for about one turn". I am quite mediocre chess player, but most of what I can play, I owe mostly to this lesson. Somehow, when developing Galaxy-brain size plans, people completely forget they are not the only actors and think everybody else is an NPC that would sit and wait until activated by their plan, and then act exactly in a way their plan needs them to. It is not so.
Making compute platforms a very rare and insanely prized resource would create an opportunity to profit by producing them. Of course, the bootleg ones would be of inferior quality, unreliable and extremely expensive. But if it would mean for India getting advantage over Pakistan - or vice versa - and not just some small advantage, but superpower-sized advantage - they'd do it, whatever it costs. They'd bribe whoever needs to be bribed, steal whatever needs to be stolen and spend as much as will need to be spent (and murder if somebody would need to be murdered, have no doubt about it). When we're in cooperation mode, of course it's much easier and cheaper to buy this stuff. But this leads to a dangerous illusion where you think that since everybody is buying the stuff from you, you're in control of everything. This is a common mistake, most recently made by one Vladimir Putin, for example.
In truth, if cooperation breaks down, there are always alternatives. Inferior, more expensive, less pleasant - but they exist, and they will be used if the preferred venue is no longer available. If the US would go to war with the world over AI, the world would learn to live without the US, and the AI will still be created - assuming, of course, current road indeed leads to AI creation. Maybe 20 years later, and maybe it will speak Chinese or Hindi - not sure how it's better - but saying "everybody will bend to my will because I have this MacGuffin" rarely works. People will find a way around, if the incentive is large enough - and if the claims of the alarmists are true, the incentive here is nearly infinite.
China tried to buy and bribe and coerce and woo people who could enable them to produce competitive hardware. They started in the cooperation era already. They dedicated over a trillion to the task. It took them decades to admit failure, and new American sanctions are mopping up their silly projects.
Iran tried to catch up to a century-old technology to secure what it perceives as its existential interests, and still tries, and fails, and will fail in the future.
The logic that the opponent will get what is necessary for him to get amounts to magical thinking. Incentives do not magically transform into actualized successes. Yes, your opponent will make moves, but if you have decisive advantage, you will still be able to thwart them.
And some «bootlegs» are bad enough to not matter. Hesbollah has pipe rockets. Israel ballistic missiles and Iron Dome. It is impossible for Hesbollah to do crippling damage to their enemies with what they have, even if they literally sacrifice all of their members for it. With next-gen AI accelerators, it's not very different. At some point in the not-too-distant future, training runs affordable to Google and Microsoft and Department of Energy will require gigawatts of power supply and a small beach's worth of silicon in conventional hardware, with no time to build either.
That's assuming there even is insubordination. China bought muh overpopulation, China recognizes climate change, China pioneered Covid overreaction, China will possibly adopt extreme Yuddist position for AI )once somebody makes a convincing enough Xinnie Pooh animation with a diffusion model). Other bad guys aren't even playing.
The world would lose. Likely will. Luckily for Americans, most parts of the world that matter will submit to their moral authority even without that Herculean undertaking.
More options
Context Copy link
More options
Context Copy link
I agree with all of this, but what do you say to those of us that think it'll be a fascist disaster, but think it might be our best hope anyways?
I'll note that on your side you have the brilliant Robin Hanson. But he also seems to be fine with handing off the future to machine descendants.
I say they should substantiate their estimates before picking the well-known evil we've been warned about so many times before over the evil they loosely extrapolate.
More options
Context Copy link
Negative Utilitarianism is just as much in play as positive. It is not a stretch to believe a boot stomping on a human face forever is a worse outcome than for a short time.
That's a fair point, although I think that argument cuts both ways.
More options
Context Copy link
More options
Context Copy link
I may also note every movement that murdered millions of people in 20th century was doing it for the sake of the future. Maybe it suggests some pattern here? Some general rule, like "if you see somebody calling for suspension of moral inhibitions today, for the sake of the bright future, you may be looking at somebody, killing whom by time travelers would be a common topic in future science fiction works". Or at least I'd be very, very cautions with such claims, just on the strength of the prior experience.
You know who else planned for the future? Hitler! I mean, I agree, but this seems like a fully general argument against planning for the future.
I agree that invoking the devil-we-know to save us from a potentially worse devil is a terrible idea unless we're very sure it's going to be worse. But I'm saying that it's likely that it will be worse. I think you're right to be skeptical, and I'm only like 60% sure myself.
No, you missed half of it. What it should have been is "you know who else called to sacrifice present morality in service of the bright future? Hitler!". Yes, I am ready to stand behind this comparison.
Oh, ok yes that is a little more specific. And I do think it's a reasonable comparison. But perhaps another reasonable comparison would have been to the Allies in that same war. I'd say both sides threw their weight behind (notionally temporary) totalitarianism and sacrificed huge amounts of value and lives in the name of the greater good. So maybe then the closest analogue to your position would have been the pacifists on both sides?
More options
Context Copy link
More options
Context Copy link
To add a point in your favor, perhaps every communist revolution ever could point to real harms of the Tsar, or capitalism, or whatever. But what they mostly got was destructive civil war, followed by grinding misery and totalitarianism.
That said, I still buy the argument that in the long run, no un-upgraded human brains will be in effective control of anything important. Robin Hanson basically says, yes that's OK. I guess I'm thinking our only hope is to slow things down enough to upgrade human brains + institutions so they can keep up.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
"I done two tours of duty in I-i-ran,
I came home with a brand new plan,
I buy GPUs from Mexico,
build a data center down Copperhead Road..."
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link