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 -
I'd like to take a moment to comment on a bit of controversy below.
With the AI safety discourse escaping containment and what might be described as the cold light of day falling across Yudkowski's forlorn-hope coalition and its personalities and practices, we are treated to an actual test of how the Rationalist movement has done at systematizing winning, particularly in the subcategory of allocating weirdness points. From where I'm sitting, it doesn't seem they've done especially well. Admittedly, I'm not the target audience, but it seems like arguing for "AI safety" you don't actually know how to secure via tools you absolutely can't be trusted to wield probably would be easier if you weren't also having to explain Aella and Slutcon and reverse-gangbang-bribery tactics and the Cow Question to generic normies at the same time.
Scott Alexander, apparently feeling the pinch, decided to put his oar in:
This is certainly one of the defenses of all time. Alexander has always been witty, and I get the feeling, possibly mistaken, that he kinda swung for the fences on this one. "Kolmogrov Complicity" and "no one actually cares about monogomy" pretty much put a cap on my ability to respect his opinions, but he's had a tough road, even if much of his own making. Let him rest.
Some have pointed out that this is Alexander doing the meme, and that discourse has continued here on the Motte, in this thread. Several commenters spoke up in support of Scott's argument, and I think it might be best to address them collectively:
To those of you who thought Scott's argument was in some way a reasonable one, I'm going to go out on a limb and say that most of you either participated in or at least had a reasonable familiarity with the Atheist internet wars. Can any of you recall a single time in that era when you heard of a prominent Conservative Christian caught in a sex scandal, where you saw it argued in good faith that actually, getting caught with a hooker was fine, because Jesus also associated with prostitutes? Can you claim to me, with a straight face, that you would consider this, then or now, a reasonable defense of a Conservative Christian influencer attending an orgy? Why do you think that is?
It's one thing to argue that Christian rules are stupid. That's understandable from non-Christians, though it's kind of neat how the logic convolutes in cases like the Cornell 7. And it's understandable for non-Christians to heap scorn on Christians who fail to live up to Christian standards, and even to then scorn all Christians for having such standards. It's not charitable, and it certainly isn't rational or principled, but humans inevitably human. When the argument is that behavior that Christians find deplorable, behavior that Christians police other Christians over, behavior that non-Christians scorn Christians for engaging in when they're caught doing so, is actually comparable to the behavior of Christ, it's time to examine the question of what further discourse is actually expected to achieve.
Scott's argument is a contemptable language game. I do not believe Scott nor anyone else is actually confused about any of the particulars here. Jesus did indeed associate with prostitutes; he did not partake of their services, nor did he employ them on their backs for horizontal forms of evangelism-by-the-hour. Christians' moral objections to Aella's activities and to the Yudkowskiites' exploitation of them for tactical advantage are straightforward and unimpeachable from within the Christian frame. There are no meaningful parallels here to the life or ministry of Jesus Christ, and implying otherwise is merely a demonstration that Scott and those agreeing with him are willing to say pretty much anything if they perceive tactical utility in doing so.
I do not accept that the Yudkowskiites or the "Effective Altruism" movement they have parasitized are offering a "message of compassion" any more than Aella or her cohort would accept that I am motivated by genuine concern for their welfare and eudaemonia, no matter how (sincerely!) I might argue that this was the case; we do not share anything remotely resembling coherent values, and it seems to me that one of the prerequisites to even attempting communication across that divide is to refrain from this sort of chicanery. My understanding of Yudkowski's legacy to date is that he has inspired at least one no-shit murder cult, a number of normal cults with a serious record of abuse toward their participants, one of the largest financial frauds ever committed, and the initiation of the exact sort of AI race he set out to avoid, together with some fairly-good short-term misery amelioration of questionable sustainability and a lot of intelligent, sensitive humans driven mad by overwhelming concern over shrimp welfare.
There's likely more to say, but why bother when people can simply pretend to not understand?
This seems to be an extremely hostile summary of EY's impact.
First off, LW+offshoots is the most interesting collection of thinkers I have encountered on the internet. A lot of people thought likewise, which is why the ratsphere enjoys some popularity. It attracted a lot of weird smart people, and obviously some of them turned murderous or financially irresponsible later on. But if Eliezer wrote a sequence called be trans and murder people, I must have missed it, likewise with the one about Forget Kelly: always bet the future of mankind at even odds.
That his attempt to argue for the danger of AI x-risk was mostly successful at arguing that higher AI capabilities are possible is true enough. Still, I would not give him credit/blame for the LLMs too much, he might have pushed capabilities ahead a few months by trying to get Musk interested in the topic, but he hardly invented the transformer architecture single-handedly.
Blaming him for "humans driven mad by overwhelming concern over shrimp welfare" is even more indirect. AI x-risk aside, EY does not seem to be very involved in EA, though obviously there is some overlap between these wider movements. At this point, you are not even blaming Jesus for medieval witch burnings, but blaming Plato for them. Also, do you have a list of these people driven "mad"? Are we talking psychosis or bipolar or some other form of "madness"? Is there even the flimsiest psychiatric study supporting your claims?
Nor do I think that the lifestyle choices of the center of LW matter much. For the longest time, Scott Alexander was to me just an entity which generated text over at LW and SSC. My respect for him is entirely based on his writing. If I had learned (before he was outed) that actually, he was in a polycule, a trans-man, a devout Hindu, into fucking donkeys, an LLM, super into NASCAR, had spent a decade in juvenile for manslaughter, that would not have mattered to me. (By contrast, if I had learned he did not have a medicine degree or actually spent most of his surplus income on torturing small mammals, that would have been disappointing.)
The reason why a priest having a mistress is generally considered a scandal is because the priest spent a lot of time denouncing that lifestyle. By contrast, the grey tribe is generally liberal in their position on sex. Aella and the polycule culture seem like a minor side note to the ratsphere than the centerpiece.
I suppose that you strongly disapprove of grey tribe sexual norms, but from where I stand, they are not affecting the credibility of LW much. If every poster on LW was married with 2.5 kids and went to church on Sunday, this would not affect Trump's stance on AI alignment at all. The Christian right would still find plenty of other reasons to hate us, just as the woke left finds plenty of reasons despite agreeing with us on most matters of sex.
Now, if the ratsphere was pushing for poly relationships as a method of alignment, you would have some reason to be concerned regarding the culture war. In fact, they do no such thing. I see Eliezer mentioned more frequently in the mainstream press these days, but his sexuality is neither pushed in the foreground by himself nor by journalists.
Aella using her area of exertise to try to help on AI alignment is not more surprising than if Scott was offering free psychotherapy to important AI researchers. Neither seems especially newsworthy.
While Scott is a blogger that writes thoughtful and occasionally entertaining blog posts, I certainly couldn't care less about where he stands in relationship to donkeys, NASCAR or juvenile justice, unless he chooses do discuss one of these topics on the blog. HOWEVER, if he comes forward as a part of a movement that claims them - or people selected by them, or people selected by the process influenced by them - need to receive extraordinary powers which will possibly control the destiny of humanity for millennia forward - and while I don't know if it's true or not, they certainly claim it is true - I think I am entitled to apply strict scrutiny here. Weird deviant blogger is not a threat to me, and I am fine with having dozens of them around without feeling the least bit unsafe. Weird deviant cult vying for absolute control over one of the most important technologies in the history of humanity - hey, that's maybe not what I am fine with. And if they proclaim they can align the AI, maybe it's time for me to ask - align AI with what exactly, and do I really want the AI to be aligned with that?
Is anyone arguing for "give Eliezer and Scott full exclusive access to prompting the first ASI instead of Sam or Dario"?
Last time I was checking, they were mostly pushing for pausing the training of bigger models. Their take is not "we have solved alignment, here is the perfect list of preferences to give AI", but "alignment seems really hard, we should not build ASI until we have solved it".
Perhaps Eliezer believes in his heart of hearts that the best outcome is a society which uses the tools of totalitarianism to crack down on any AI development, but he is not in the position to do that and will never be.
If we manage to coordinate and build aligned ASI, then we will not simply put Eliezer's value system into it, because quite frankly Eliezer is not a critical stakeholder with whom one needs to coordinate. Likely the USG would get 30%, Anthropic and OpenAI would get 10% each, the Chinese would get another 30%, the rest of the world would get another 10% and LW would end up with less sway than the Catholic Church.
It's not personally Scott and Eliezer. It's the Oversight Committee (there's always a committee). But who do you think will influence forming the committee and who do you think the members of the committee would look for guidance when faced with tough decisions? Wouldn't it be the folks that convinced the government that the committee is needed in the first place? Or at least folks aligned with them?
He personally may be not. But some more savvy, slick and energetic politician taking cues from him might be. Marx or Gramsci did not personally take power.
If Eliezer does not matter, then it doesn't matter who says what about him. I am fine with this outcome, where whatever he says and does is immaterial and he has no influence on the outcome of things. I am afraid however it's not entirely the case, as a lot of people who are very knowledgeable in the industry and are at least adjacent to its control levers are significantly influenced by him (either personally or by the ideology he embodies). Maybe I am worrying for nothing, I'd be happy with that outcome too.
More options
Context Copy link
"We should not build AI to..." involves putting pressure on companies and legislators, and can easily lead to a situation where users don't get to control their own computers. "Ai should refuse to do X" is ultimately an attempt to DRM AI in the same way that media and video games are already DRMed.
If the experience with gun control or drug control teaches us something, it's how hard is to prevent people from doing something they want to do. And how much control does it require to have even a limited degree of success in it. Imagine a world where only ISIS and the Sinaloa cartel has the true AI... And the CIA/NSA of course but we don't talk about it. OTOH the regular citizens ae required to have government monitoring software on any device smarter than a thermostat.
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
More options
Context Copy link