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 -
Apparently, a lab in china has created a virus with a 100% kill rate in humanized mice. Combined with the fact that there's a decent chance that COVID was a lab leak, this sort of thing is extremely dangerous to be doing.
I'm not sure how best to make it so that people are not incentivized to do things like this, but ceasing to fund this variety of research (it looks like the US ended one program that was pushing this sort of thing last year), and instating some sort of legal liability on those who do this, and especially if they dispose of it badly, probably seem like good decisions.
Extremely dangerous diseases are among the top few things in being both disastrous to humanity (unlike climate change) and also relatively likely (unlike a massive asteroid hitting earth). Development of them is also something that is not excessively difficult to do. This is probably the closest thing we have so far to Bostrom's black ball metaphor. People joke about Yudkowskian airstrikes on data centers; would airstrikes on labs be similarly warranted? More seriously, though, there should be far more effort put into preventing this sort of thing than there currently is.
Bostrom's concerns should probably be something more important to be aware of. The ideal is just to not develop technology in specific fields to the point that killing millions is a cheap and easy thing to do. Of course, the tradeoff is totalitarianism, a terror of its own.
EDIT: Some of the comments have argued, relatively convincingly, that this particular news story was overblown and misleading.
China has nukes. Everything Chinese labs do is (to some extent) state sanctioned. What choice do you have? This was always the most idiotic thing about Yudkowskian airstrikes, given the growing technological adeptness of America’s foes. Why would one risk a certain chance of total nuclear war for a partial chance of a killer virus that - at worst - would kill me just the same? It’s poor logic.
The only reason to pick nuking datacenters or virology labs over not doing so would be if your commitment to mankind was so great that you would accept your certain death in exchange for some South Americans and Africans surviving global nuclear war and repopulating the earth in thousands of years.
Alas, I am not that selfless.
He wanted it to be multilateral, where the big powers agree to force the little powers into line. That's the stupidest part of it, the idea of multilateral, genuine, sincere enforcement of a rule as opposed to 'AI for me, not for thee' like 'H-bombs for me, not for thee'.
Why is that stupid?
Because the great powers are incapable of cooperating in this unselfish way. Nuclear arms control is my example - the whole idea started just after the big powers acquired their nuclear arsenal and only applies to weaker latecomers.
Yudkowsky is like those who want global nuclear abolition, where neither big nor little powers have nuclear weapons but for AGI research. No country is going to consciously and deliberately kneecap their capabilities and fall behind in the race, especially in times like these when a competitive edge is in high demand. And AI is even more hard to ban than nuclear weapons. All the strongest lobby groups want more AI and AGI - big tech, big corporations, militaries, state security forces. There's no strong lobby group against AGI like there is against nuclear arms races, the risks are less obvious. AI is profitable and provides economic dividends, unlike piling up huge numbers of nuclear weapons. AI is so much harder to ban than nuclear weapons and we can't even do the latter.
Furthermore, the only two countries with a chance at AGI are the US and China, they're opposing forces. Of course they want to get ahead of the other, that's Made in China 2025 and the US CHIPS act in a nutshell.
Yudkowsky wasn't saying "this is likely to happen." He was saying "this is the sort of arrangement where humanity could avoid being made extinct by a hostile superintelligence." Which seems right to me if you agree with the premise that superintelligence is dangerous and possible and hard to align.
The difference between nuclear weapons and ASI is that ASI kills everyone and nuclear weapons don't. If people realized that then it would not be hard to ban it. Imagine if one nuclear weapon destroyed the whole solar system. Do you see how a treaty banning them would not be difficult? Even if China wasn't convinced, it still would not be that difficult to convince or prevent them from building one. Far easier than WWII, as Yudkowsky has said.
More options
Context Copy link
A notable example of how this works is the war in Ukraine. Ukraine was coaxed into not being a nuclear power post-USSR. At the time, this definitely seemed like a good idea. Now they are in a situation where they and the West are always afraid of escalation, because of Russia's nuclear weapons.
(The US has a track record getting on the wrong side of escalation: in the Vietnam War, the US held back from e.g. a naval blockade of Vietnam, to avoid China and the USSR becoming more involved. In the Korean War, the US didn't use its nuclear arsenal, long before MAD, to avoid escalation. I'm not saying that either decision was wrong.)
More options
Context Copy link
Nonproliferation maybe, but big powers have done lots of arms reduction and test ban treaties.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
Well yes, he said he wanted it to be multilateral but did (as I recall, I might be misremembering Twitter posts) suggest hostile action toward ‘rogue’ international actors (including well-armed or major ones) would be justified. And yes, it doesn’t seem likely you’d get a global agreement with zero defection, and if you did you’d have solved the major impediment to world peace.
You may well be right, I don't have a good memory of what he said either (or how he later tried to clarify/sanewash). Feels bizarre to spend so much time thinking about which parts of an unfeasible idea are most unrealistic, given it's clearly dead in the water now. US chip sanctions have cemented a race dynamic. That Washington doesn't want Beijing to have these things will surely make them even more enticing and poison trust.
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