Daniel Drucker

Why do we derogate effective altruists, activists, & other radically prosocial individuals? In new work, we discuss how doing good that deviates from social norms gets stigmatized. New preprint w/ @dcameron.bsky.social @tlau.bsky.social @desmond-ong.bsky.social: osf.io/preprints/ps...

Would you double up on "yabba"s in the full phrase or just go with "yabba doo"?

Fine, I'm not sure that's wrong about guilt-by-association arguments generally, but the associations are extremely weak in this specific case.

Maybe short-term? Long-term it is not good or rational politics that benefits from those modes of engagement, in my opinion.

That sounds a little like justifying dishonesty, though?

Guilt by association arguments are often weak even with strong associations, which is very much not the case here. Argue against the idea itself!

One thing that's fairly crazy-making to me is how rarely people are suspicious that the thing they want politicians to run on are the things that would be most emotionally satisfying to them for the politicians to run on. It's great when they align, but you should be worried you're forcing it.

Endnotes for a start. Two-column portrait doesn’t work great for reading philosophy imo, but I think that may be more idiosyncratic.

They brought back endnotes though!

Why did PPR ruin its layout so badly?? philpapers.org/go.pl?aid=RO... (the paper looks interesting, no shade to the paper)

Amazing typo

SALT 36 will be held at my alma mater, the University of Buenos Aires, on July 29-31 2026. This will be the first time the conference takes place in South America.

Abstract deadline: Dec 15, 2025
Link: saltconf.github.io/salt36/

When a post begins by skeptically invoking my training as a philosopher I know I'm in for a good time

Don't threaten me with a good time etc etc

(I don't mean for it to be a universal norm for all posts, but a lot of stuff has had a "THEY don't want you to know about x" feel for me lately.)

Right! I mean as a general phenomenon we'd all be better off giving specific quotes, because attribution standards seem problematically loose these days.

Doesn't have to be here, and it's not really specific to this at all, but I've been seeing so many misattributions of views to Klein lately that I think people in general would be better off giving specific quotes when available.

Is there anyone who says there's no upside to losing the battle? I think what they'd say is that in most cases where someone's defending losing the battle for coalition inspiration, the consequences of losing the election are dire enough that the hope for greater strength later is not worth it.

This is my regular reminder to everyone that jstor is open to the general public now; a free account there will give you access to 100 papers a year.

There's some good philosophical/academic/scientific discussion on here, since it's where most of the academics are now I think. I'd say it takes roughly as much curation as Twitter as far as who you follow, but the bystanders are better, by and large, once it's not mostly politics.

I promise there are islands of sanity. My feed's average quality is much higher than on Twitter's. (The kind of politics Itai was wading into does bring out this place's worst characteristics, tjough.)

Right, I think that's part of it, but then there's a related one that it's easy to confuse it with, namely whether they're subject to rational norms, etc. I think it may be a consequence of satisfying the second criteria, but may not be, and I think people care about it independently.

I suspect 2 is the crucial one, and it needs so much more refinement (what is comprehension?) that all the same worries will arise again.

Yeah, since ad hocness is essentially a question of motivation, our motivations are most evident given our priors.

It’s the easiest way. Another way to mitigate the feeling of ad hocness is to give good theoretical grounding of the changes. But if it changes too much in close proximity to the realization that the criteria count LLMs as intelligent, it will still seem objectionable as hoc.

If the criteria change frequently not in response to actual prior counterexamples (parrots say) but simply when the theorist recognizes that, by those criteria, LLMs are intelligent.

Given this, we agree, so long as you don’t do it too much, especially in ways motivated just by classing LLMs themselves as unintelligent.

Genuinely confused how you got that from what I said? I’m saying that in inquiries where what you care about is finding stuff out, messing with criteria constantly in order to avoid categorizing apparent like with like is not typically reasonable but more like playing a no-lose game.

(And inquiries guided by our desires in this way are rarely successful or reasonable, we know from long experience.)

If your criteria constantly change in ad hoc ways to rule out the intelligence of some system your prior criteria would've included, you look less guided by a conception of intelligence that you're articulating sub-optimally, and more guided by the desire to ensure these systems are unintelligent.

Wouldn't being self-undermining be more, "here's some specifiable alternative superior way of doing things I could adopt, but I'll stick with my way"? Thinking your ways are generically superior goes beyond that, plausibly irrationally I think?