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 -
To expand on a comment I made in the previous thread, the situation in Minnesota has made me realize that in the impending AI revolution, in the best case scenario (where we don't get slaughtered by Skynet; we have machines to do our labor for us; and there is some sort of social assistance for all the newly unemployed) it's pretty likely there will be a huge amount of social unrest, separate and apart from issues arising from the AI revolution.
I had assumed that in such a scenario, most people would either pursue interesting hobbies or self-destructive hedonism. But the situation in Minnesota makes me realize that a lot of people are going to look for political causes and use those as an excuse to harass others while feeling morally superior in doing so.
Obviously these things are very difficult to predict, but the Summer of Floyd is instructive, I think. A lot of the rioters were people who were furloughed during Covid and collecting unemployment insurance.
You are also not contemplating the most likely middle-ground nightmare scenario: There is no singularity, but AI is good enough that it puts most of the middle class out of work, there is no UBI and now you have a bunch of people who are purposeless, humiliated, have a lot of free time, and are pissed off and have nothing to lose due to their now degraded economic and social state.
Or the converse: AI gets just strong enough to keep the resulting bunch of purposeless, humiliated humans under control.
Yeah. These middle-ground scenarios are so absurdly under-discussed that I can't help but see the entire field of AI-safety as a complete clownshow. It doesn't even take a lot of imagination to outline them.
Middle ground plateaus aren't particularly likely and anyone who thinks about the problem for more than it takes to write snarky comment should understand that. In any world where AI is good enough to replace all or most work then it can be put towards the task of improving AI. With an arbitrarily large amount of intelligence deployed to this end then unless there is something spooky going on in the human brain then we should expect rapid and recursive improvement. There just isn't a stable equilibrium there.
Alignment is about existential risk, we don't need a special new branch of philosophy and ethics to discuss labor automation, this is a conversation that has been going on since before Marx and alignment people cannot hope to add anything useful to it. People can, should be, and are starting to have these conversations just fine without them.
I don't think "the government decides to pump the breaks on AI development after it becomes powerful enough to control the populace but before it becomes to powerful to be controlled" is a particularly unlikely outcome (accepting for the sake of argument that such an uber-powerful AI is possible).
More options
Context Copy link
Yeah, "make a completely unsubstantiated statement in order to justify a singular focus on fanciful scenarios that you don't know will ever take place" is exactly the sort of thing that prevents me from taking the field seriously.
The issue isn't even labor automation, it's things like "we now have technology that makes the world of 1984 possible", and we're already there without even reaching full labor automation. It's just a question of building out infrastructure, and this isn't even one of the more imaginative scenarios.
Calling it non-existential is cope. As a threat it's far more likely, and we have zero counter-measures for it. Focusing on scenarios that we don't even know are possible over ones we know are possible, and we are visibly heading towards them, is exactly my criticism.
Your complaint appears to be that this group of people concerned specific with a singularity event needs to instead focus their efforts on something you don't even seem to think AI is needed to make happen. And as an aside, all the thinkers I've read that you would consider AI-Safety aligned have in fact voiced concerns about things like turning drones over to AI. Their most famous proponent, big yud wants to nuke the AI datacenters.
You're just describing a subset of unaligned AI where the AI is aligned with a despot rather than totally unaligned. Or, if the general intelligence isn't necessary for this, then it's a bog standard anti-surveillance stance that isn't related to AI-safety. The AI-Safety contingent would absolutely say that this is an unaligned use of AI and would further go on to say that if the AI was sufficiently strong it would be unaligned to its master and turn against their interests too. The goal of AI safety is the impossibly difficult task of either preventing a strong AI future at all or engineering an AI aligned with human interests that would not go along with the whole 1984 plan.
Huh? No, AI is necessary to make it happen, but the current version that we have is sufficient. Like you point out, it would make no sense for me to bring it up in an AI conversation otherwise.
Yes, because he's obsessed with fantasy doomsday scenarios, rather than far more realistic ones. That's my criticism.
Everything I saw from the rat-sphere of the subject, including the concept of "alignment", assumes AI will have agency, and goals that it will be pursuing. None of it is necessary for the dangers that AI will bring.
Again, defining the field in such a way that it ignores the most likely risks, is exactly the issue I have with AI-safety.
How is that useful? I don't care about what they call "aligned" and "not aligned", I care about how a given scenario could come about, and how it could be prevented (and no, "nuke data centers" doesn't count). This would be another part of the criticism I have of the entire field.
The "AI-Safety" people as you call them have a particular interest in alignment as AI hits super intelligence. They don't need to be wearing their "AI-safety" hats to oppose a surveillance state. You don't need any kind of special MIRI knowledge to oppose surveillance states and people have opposed them for a long time. This is the kind of scope creep criticism that leftists do when the accuse climate focused causes of not focusing enough on police injustice against BIPOCs.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
...Or unless intelligence suffers from diminishing returns, which actually seems fairly likely.
Where do these diminishing returns kick in? Just within the human form factor we support intelligences between your average fool and real geniuses. It seems awfully unlikely that the returns diminish sharply at the top end of the curve built by natural selection under many constraints. Or maybe you mean to application of intelligence, in which case I'd say just within our current constraints it has given us the nuclear bomb, it can manufacture pandemics, it can penetrate and shut down important technical infrastructure. If there are some diminishing returns to its application how confident are you that the wonders between where we are now and where it diminishes are lesser to normal distributional inequality that we've dealt with for thousands of years?
Within the human scale, at the point where Von Neumann was a functionary, where neither New Soviet Man nor the Thousand Year Reich arrived, where Technocracy is a bad joke, and where Sherlock Holmes has never existed, even in the aggregate.
We can do all those things. Can it generate airborne nano factories whose product causes all humans to drop dead within the same second? I'm skeptical.
Did a notably finite number of very smart people produce nuclear bombs yes or no? Can a notably finite number of very smart people almost certainly produce a super pandemic yes or no? And these are the absolutely mundane appliations of intelligence.
It seems to me that there is a long tradition of smart people coming together an inventing new and not distantly in the past foreseen weapons and technologies. The very nature of these advancements not being seen far before they came about makes conjuring up specific predictions impossible. You can always call anything specific science fiction, but nuclear was science fiction at one point. And there is of course just the more mundane issue of a sufficiently advanced AI that is merely willing to give cranks the already known ability to manufacture super weapons could be existential.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
Does it? The human brain is only about three times larger than the chimpanzee brain. But that 3x difference enabled us to take over the world. Or, as Scott put it:
I've got to say, sometimes it is pretty funny being on a board where two of the abiding topics of concern are, distilled down a bit, "high IQ people being wiped out by lower-IQ people" and "high IQ AI wiping out lower-IQ people."
Anyway, there's obviously not a direct correlation between intelligence and existential risk. Creatures with an IQ of 0 on a scale of 1 - 100 for intelligence are in a far less precarious position, existentially speaking, than creatures with an IQ of 100 (us). Intelligence is only an imperfect proxy measurement for power and power is what generates existential risk.
More options
Context Copy link
It seems to me that it does, yes. If your intelligence scales a hundred-fold, but the complexity of the thing you want to do scales a billion-fold, you have lost progress, not gained it. The AI risk model is that intelligence scales faster than complexity and that hard limits don't exist; it's not actually clear that this is the case, and the general stagnation of scientific progress gives some evidence that the opposite is the case. It seems entirely possible to me that even a superintelligent AI runs into hard limits before it begins devouring the stars.
Now on the one hand, this doesn't seem like something I'd want to gamble on. On the other hand, it's obviously not my choice whether we gamble on it or not; AI safety has pretty clearly failed by its own standards, there is no particular reason to believe that "safe" AI is a thing that can even potentially exist, and we are going to shoot for AGI anyway. What will happen will happen. The question is, how should AI doomsday worries effect my own decisions? And the answer, it seems to me, is that I should proceed from the assumption that AI doomsday won't happen, because that's the branch where my decisions matter to any significant degree. I can solve neither AI doomsday nor metastable vacuum decay. Better to worry about the problems I can solve.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
I can’t rebut the statement that these scenarios are “under-discussed” (since that depends on how much you personally think they should be discussed), but it’s certainly false that these scenarios go undiscussed. For examples of long essays written about (approximately) these scenarios, see here and here.
From the intro to the former essay:
Seems pretty similar to what’s being discussed in this subthread.
It's not just automation, the discussed scenario was " AI gets just strong enough to keep the resulting bunch of purposeless, humiliated humans under control". My emphasis would be on the "under control" part. Even when discussing automation, they have tendency of veering off into fantasy scenarios of full-automation, when the more likely ones are comparative-advantage mediated push towards menial labor in service of the AI god.
I mean, reading these essays, they seem to be pretty focused on the idea that AI can be used for mass control and suppression. Chapter 4 of the first essay that I linked is rather explicit about this:
Is your contention that these discussions are predicated on “full automation” scenarios while you think that there aren’t any obstacles stopping an AI-powered tyranny from happening now?
Sort of. My contention still boils down to "under-discussed", that the issues that are more likely to happen take up less focus than ones that are less likely to happen. The "full automation" thing is an example of this - AI developing to the point where it replaces literally everyone / the vast majority of people can happen somewhere down the line, but a scenario where everybody still has a job, because it makes more sense to let AI specialize in data professing, while humans focus on menial jobs is more likely, and unpleasant enough to warrant discussion.
I only had a skim of the essay you linked, and it's indeed more like what I'd like to see, but not quite there yet.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
I mean, it doesn't take the lack of discussion regarding middle-ground scenarios to see the entire field of AI-safety as a complete clownshow.
But it doesn't hurt, either.
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