site banner

Culture War Roundup for the week of August 11, 2025

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.

7
Jump in the discussion.

No email address required.

Training language models to be warm and empathetic makes them less reliable and more sycophantic:

Artificial intelligence (AI) developers are increasingly building language models with warm and empathetic personas that millions of people now use for advice, therapy, and companionship. Here, we show how this creates a significant trade-off: optimizing language models for warmth undermines their reliability, especially when users express vulnerability. We conducted controlled experiments on five language models of varying sizes and architectures, training them to produce warmer, more empathetic responses, then evaluating them on safety-critical tasks. Warm models showed substantially higher error rates (+10 to +30 percentage points) than their original counterparts, promoting conspiracy theories, providing incorrect factual information, and offering problematic medical advice. They were also significantly more likely to validate incorrect user beliefs, particularly when user messages expressed sadness. Importantly, these effects were consistent across different model architectures, and occurred despite preserved performance on standard benchmarks, revealing systematic risks that current evaluation practices may fail to detect. As human-like AI systems are deployed at an unprecedented scale, our findings indicate a need to rethink how we develop and oversee these systems that are reshaping human relationships and social interaction.

Assuming that the results reported in the paper are accurate and that they do generalize across model architectures with some regularity, it seems to me that there are two stances you can take regarding this phenomenon; you can either view it as an "easy problem" or a "hard problem":

  • The "easy problem" view: This is essentially just an artifact of the specific fine-tuning method that the authors used. It should not be an insurmountable task to come up with a training method that tells the LLM to maximize warmth and empathy, but without sacrificing honesty and rigor. Just tell the LLM to optimize for both and we'll be fine.

  • The "hard problem" view: This phenomenon is perhaps indicative of a more fundamental tradeoff in the design space of possible minds. Perhaps there is something intrinsic to the fact that, as a mind devotes more attention to "humane concerns" and "social reasoning", there tends to be a concomitant sacrifice of attention to matters of effectiveness and pure rigor. This is not to say that there are no minds that successfully optimize for both; only that they are noticeably more uncommon, relative to the total space of all possibilities. If this view is correct, it could be troublesome for alignment research. Beyond mere orthogonality, raw intellect and effectiveness (and most AI boosters want a hypothetical ASI to be highly effective at realizing its concrete visions in the external world) might actually be negatively correlated with empathy.

One HN comment on the paper read as follows:

A few months ago I asked GPT for a prompt to make it more truthful and logical. The prompt it came up with included the clause "never use friendly or encouraging language"

which is quite fascinating!

EDIT: Funny how many topics this fractured off into, seems notable even by TheMotte standards...

You know, I've long noticed a human version of this tension that I've been really curious about.

Different communities have different norms, of course. This isn't news. But I've had, at points, one foot in creative communities where artists or crafts people try to get good at things, and another foot in academic communities where academics try to "understand the world", or "critique society and power", or "understand math / economics / whatever". And what I've noticed, at least in my time in such communities, is that the creator spaces if they're functional at all (and not all are) tend to be a lot more positive and validating. A lot of the academic communities are much more demoralizing.

I'm sure some of that is that the creative spaces I'm thinking of tend to be more opt-in. Back in the day, no one was pointing a gun at anyone's head to participate in the Quake community, say. Same thing for people trying to make digital art in Photoshop, or musicians participating in video game remix communities, or people making indie browser games and looking for morale boosts from their peers. Whereas people participating in academic communities often are part of a more formalized system that where they have to be there, even if they're burned out, even if they stop believing in what they're working on, or even if they think it's likely that they have no future. So that's a very real difference.

But I've also long speculated that there's something more fundamental at play, like... I don't know, that everyone trying to improve in those functional creator spaces understands the incredibly vulnerable position people put themselves in when they take the initiative to create something and put themselves out there. And everyone has to start somewhere. It's a process for everyone. Demoralization is real. And everyone is trying to improve all the time, and there's just too much to know and master. There's a real balance between maintaining the standards of a community and maintaining the morale of individual members of a community - you do need enough high quality not to run off people who have actually mastered some things. And yet there really is very little to be gained by ripping bad work to shreds, in the usual case.

But in the academic communities, public critique is often treated as having a much higher status. It's a sign that a field is valuable, and it's a way of weeding "bad" work out of a field to maintain high standards and thus the value of the field in question. And it's a way to assert zero sum status over other high status people, too. But more, because of all of this, it really just becomes a kind of habit. Finding the flaws in work just becomes what you do, or at least that was the case for many of the academic fields I was familiar with (I've worked at universities and have a lot of professor friends). And it's not even really viewed as personal most of the time (although it can be). It's just sort of a way of navigating the world. It reminds me of the old Onion article about the grad student deconstructing a Mexican food menu.

The thing is, on paper, you might well find that the first style of forum does end up validating people for their crappy mistakes. I wouldn't be surprised if that were true. But it's also true that people exist through time. And tacit knowledge is real and not trivially shared or captured, either. I feel like there's a more complicated tradeoff lurking in the background here.

Recently I've been using AI (Gemini Pro 2.5 and Claude Sonnet 4.1) to work through a bunch of quite complicated math question I have. And yeah, they spend a lot of time glazing me (especially Gemini). And I definitely have to engage in a lot of preemptive self-criticism and skepticism to guard against that, and to be wary of what they say. And both models do get things wrong some time. But I've gotten to ask a lot of really in-depth questions, and its proven to be really useful. Meanwhile, I went back to some of the various stackexchange sites recently after doing this, and... yet, tedious prickly dickishness. It's still there. I know those communities have, in aggregate, all sorts of smart people. I've gotten value from the site. But the comparison of the experience between the two is night and day, in exactly the same pattern as I just described above, and I'm obviously getting vastly more value from the AI currently.

And what I've noticed, at least in my time in such communities, is that the creator spaces if they're functional at all (and not all are) tend to be a lot more positive and validating. A lot of the academic communities are much more demoralizing.

I think that's probably true as a general trend, but it also heavily depends on context. A lot of art communities (writing, music, photography, etc) can be vicious, especially when there's a palpable sense that you have a lot of people competing over very few economic opportunities. And in some academic departments like English or any type of Studies department, glazing the work of others (especially the work of your direct superiors in the social hierarchy) is the norm.

And in some academic departments like English or any type of Studies department, glazing the work of others (especially the work of your direct superiors in the social hierarchy) is the norm.

Well, a very close acquaintance of mine is in an English department, and all I can say after the last 10 years is that, while there absolutely is a lot of that style of glazing (a lot of the communication styles are heavily female and rely on huge amounts of validation, or at least that's my impression), it has been tangled up with the most awful Campus Reform-style it'd-be-a-caricature-if-you-didn't-see-it-first-hand race/gender/sexuality crabs in a barrel dynamics and hierarchy arson you could imagine... and she has peers in a number of peer departments at other universities who went down that road as well. It seems like it's quieted down over the last year or so, but it was honestly beyond parody for a few years there. A whole lot of mid-career Gen X people were just putting their heads down, taking their beatings, and waiting for it all to blow over. But yes, to be fair, it actually had a deep family resemble to some of the insane art community dynamics you are describing, too, which I have read stories about.