danieldrucker.bsky.social
Philosophy professor at UT Austin who thinks about attitudes, epistemology, and communication. https://www.danieldrucker.info/
1,029 posts
783 followers
926 following
Active Commenter
comment in response to
post
Notice p and q are incompatible (if I get a puppy soon, then I'm not in a simulation, and and I'm in a simulation, I'm not getting a puppy soon).
comment in response to
post
So let p = "I get a puppy soon"; q = "I live in a simulation where everything is geared toward teaching me and giving me incredible understanding". I think it makes sense to want both of these, but it's bizarre to say "I want to get a puppy soon or live in a simulation where everything is...".
comment in response to
post
Excellent, thanks. I think we can still make the counterexamples work, if you share my intuition about the cases. That's because incompatibility isn't the kind of relatedness that'd make sense of why a disjunction might be desirable.
comment in response to
post
(Or where this is one of the few that might, who knows.)
comment in response to
post
Yeah, that's what I had in mind. It's a potential entailment I'm kind of fascinated by, and I was wondering if there's a way of having a kind of middle-ground view where most potential content-entailments don't entail belief-entailments, but where only this one might.
comment in response to
post
Another thing: why wouldn't your story apply equally well to "Bel(p and q)" not entailing "Bel(p)"? That entailment _might_ work because believing p might really be the only rational thing for an agent who believes p and q.
comment in response to
post
Ah, okay, I gotcha. So there's a bit of a dialectical reason for not going this way. It leaves part of the data unaccounted for, namely, that "it's frustrating that p" exhibits exactly the same patterns that "I'm frustrated..." does. So a story connecting them feels hard to avoid?
comment in response to
post
I'm sure this is my fault, but I'm not quite grokking your proposal. What do you mean when you propose making the criteria dependent on individual attitude states, and how is it more intensional than what I say? Could you elaborate a bit?
comment in response to
post
"berenjena_1, not berenjena_2"
comment in response to
post
quod.lib.umich.edu/e/ergo/12405...
comment in response to
post
Jeremy Fantl has an interesting paper about I think the structural analogue for happiness, where when it's initially reported that Gore won the presidency, a supporter seemed to be happy that he had won, even though he hadn't won. He thinks this really does go against factivity; I'm sympathetic.
comment in response to
post
I offer a solution in the paper, but this is already a long thread, so I'll leave it here. But I'd be delighted if other people had other ways to resolve this puzzle; it's an interesting one I've been thinking about for a while!
comment in response to
post
"It's frustrating that Cormac McCarthy died before I got to meet him or dogs live very short lives" does not sound true. In other words, frustratingness, unlike truth, does not seem to obey the following schema: "Fp & Fq, so F(p or q)". There's a puzzle here, to explain why that should happen.
comment in response to
post
Consider: "I am frustrated that Cormac McCarthy died before I got to meet him or dogs live very short lives." This sounds bad, too, but notice that unlike "it's true that Trump is president now or the atmosphere is mostly composed of nitrogen", which sounds true ...
comment in response to
post
E.g., "Trump is president now or the atmosphere is mostly composed of nitrogen." The other properties, frustratingness, desirableness, admirableness, and so on — the properties that play the relevant truth role, but for the other attitudes — work differently from truth.
comment in response to
post
I argue this explains why, though they sound strange to report, beliefs with very unrelated disjunctive contents can be held correctly and, because of that, reasonably. (If your evidence sufficiently supports the truth, etc. of p, no defeaters, etc., then it's reasonable to believe, etc. that p).
comment in response to
post
When it's true that p, your belief that p is thereby correct, or "fitting". When it's frustrating that p, your frustration that p is thereby fitting. When it is or would be desirable (maybe good) that p, desiring that p is thereby fitting. Etc. for admiration and admirableness, and so on.
comment in response to
post
Perhaps one thing to experiment with is recapturing some of the benefits of slow reading in other media that you can get people to want to consume. (Video games, if it's not too cringe to say, e.g.) Maybe some things can't be recreated that way, but it'd be interesting to explore the potential.
comment in response to
post
And... you're surprised by this?
comment in response to
post
And would you say, as the student's essay is described, that likely falls in this category?
comment in response to
post
Do you think it happens that some substantive conclusions have so much initial evidence against them that, even without seeing the particular piece arguing in their favor, you know the paper could not be meritorious? For example, if the author had argued the British wrote the Constitution for us?
comment in response to
post
My experience (in philosophy in particular) is that what the influential philosophers said is usually more interesting and meritorious than what the tradition took from them, so I guess I have the opposite take to yours.
comment in response to
post
Next installment of the Danielle Allen and Yarvin series should be Korsgaard and Thiel.
comment in response to
post
In not answering you prove your superiority over the machines.
comment in response to
post
Would you say the same thing about, say, fundamental math, out of curiosity, to the extent you have a grip on the corresponding issues there? (I very much agree with your overarching point.)
comment in response to
post
No, I don’t think so, because Lucy might not be displeased? I’m displeased that eg dogs live short lives, but I’m not displeased. We could try to say I’m “pro tanto displeased” but I don’t know what that really means.
comment in response to
post
Is there some reason not to think that both things happen? A lot of ‘seems’ and ‘intuitive’ talk is just clarifying premises, some is more mentally substantial? It’d defang Setiya’s response a bit, but I like a companion in guilt response invoking math etc more than just saying they’re premises.
comment in response to
post
I think I would consider that they would have a direct experience of a lot of things that we refer to: specifically, aspects of conversations, conversational participants, language, etc. Doesn't that give them a lot to go off of, even without relations to material things other than we humans?
comment in response to
post
Why can’t they refer by deferring to how we refer, when it’s not stuff they have contact with (tokens, us)?
comment in response to
post
Different users all have the same blob (or many do, given that Claude 4 Opus is a different blob from GPT 4 or whatever), if I get your meaning.
comment in response to
post
(And that'd be so even if the Daniels and I are genetically very similar and in education as well.)
comment in response to
post
To me, just treating them like each branch is a one-off allows enough stability for pragmatics. Many people look at them, to my mind erroneously, as representatives of a Claude, etc.-blob, but that should matter no more than if people thought I was the representative of a blob of all the Daniels.
comment in response to
post
I can't imagine being interested in LLMs but not having watched Severance, that's fascinating.
comment in response to
post
I think of it as that, if they have mental states, they have them relative to ongoing "context windows" (specific branches of them in conversation with, typically, one human user only). philpapers.org/rec/GOLLCN Simon Goldstein considers this, too, but there's more to be said about it.
comment in response to
post
Is there a reason the textual inputs and outputs don't make for a distinctive enough "environment"? Beliefs and preferences all concern streams of text they have encountered or might encounter, at a minimum.
comment in response to
post
Can you elaborate on why it's obviously so inappropriate for LLMs? I'm interested in why people take that to be obvious.
comment in response to
post
"If our official theories disagree with what we cannot help thinking outside the philosophy room, then no real equilibrium has been reached. Unless we are doubleplusgood doublethinkers, it will not last." (This continues the earlier quote.) Sophisticated I think are products of philosophy.
comment in response to
post
He doesn't give a theory distinguishing these, so I can't be completely certain. But the quote makes it very clear, to me at least, that he doesn't privilege commonsensical opinions, since they only mark a set of those that must be balanced. He doesn't mark this with pre-/post-theoretical, true.
comment in response to
post
I'm confused myself what the disconnect is... he specifically contrasts commonsensical and sophisticated opinions. Presumably some come from general, common, commonly perceived reliable opinions, and some from more abstract reasoning or commitments apart from that? What else would the contrast be?
comment in response to
post
But that’s not what he’s saying: opinions can be sophisticated rather than pretheoretical, and they must be balanced just like the pretheoretical ones.
comment in response to
post
But they are all opinions, and a reasonable goal for a philosopher is to bring them into equilibrium." I guess I think he doesn't privilege common sense? (This is from his first collection of papers.)
comment in response to
post
"Our 'intuitions' are simply opinions; our philosophical theories are the same. Some are commonsensical, some are sophisticated; some are particular, some general; some are more firmly held, some less. ...
comment in response to
post
(Sorry for the tortured syntax in the first sentence :X)
comment in response to
post
This doesn't go very far as far as actual conservativeness goes, though, right? Lewis believed in talking donkeys, e.g.; even looking from his metaphysics, he kind of argued that mainstream Christians are akin to devil worshippers, among other things.
comment in response to
post
“Pick one” is the curse of liberal politics in the US post-Obama.
comment in response to
post
Good of you to allow some people to become fathers ;)
comment in response to
post
To clarify: my suspicion is that the prominent proponents of the thesis did not mean for it to entail anything about power, and the objectors make it about actual power to avoid discussing ideology/ambitions.
comment in response to
post
Oh, I took him to deny he was a fascist. I meant proponents of the fascism view.
comment in response to
post
Is Setiya’s description of the project accurate: “unmasking the hidden influence of liberal dogma on the scope and methods of ‘analytic philosophy’”? And if so, did he make the case for that well? Is there an argument for it that Setiya fails to address that you find compelling?