[onecornerface]

Give me money (if you want)

Okay I have a way to receive money. Click the part that is underlined.

I do not make much money, but I am not desperately in need. Consider whether your money will be spent better elsewhere. I have a family to rely on. But I would like to rely on them somewhat less.

If you like some of my posts, pls consider giving me a small one-time sum of money. I will give you no rewards except my appreciation. Thank.

Here is the link again

[onecornerface]

I hate when I make a very long and detailed post, where I am careful to use specific phrasing, provide relevant caveats, acknowledge limitations, cover (at least in brief) a wide range of important closely-related issues, etc. --and nobody reads it, and it gets zero or extremely few likes/comments/reblogs. Then I make a crude and oversimplified short post on the same topic, and several people reblog or comment to say it contains distortions, sloppy phrasing, lack of hedging qualifiers, or omission of significant issues.

Well yeah, it was a short post, so I think it will almost inevitably have these problems. A post where I put great effort into avoiding such problems almost invariably becomes a longer post-- which then you very likely wouldn't have seen or responded to.

In my longer posts, I think I do address these kinds of issues OR at least demonstrate that these are the sorts of things that I indeed do think about a lot and that I am capable of addressing these issues at painstaking length.

At the end of the day, the posts are what they are, they appear isolated in the feed for various people, and the context (including of my other longer posts) is not self-evident-- and so it is certainly fair game to criticize a short and very flawed post that I make, even if you see it out of context. But I hate how often I seem to be judged by my short flawed posts, which just do not reflect the nuance and effort and detail of my actual thinking as seen in my longer higher-quality posts that do not get read.

[onecornerface]

Fatal opioid overdoses have notably decreased in the US and Canada since 2024.

If naloxone is a main cause of the decrease in fatal overdose, as is often claimed, then shouldn't we see an increase in nonfatal overdose alongside the decrease in fatal overdose? But there appears not to have been an increase in nonfatal overdose.

I almost never hear anyone trying to dig into these numbers and actually figure things out. These are basic questions that should be common sense. Maddening.

On average, prohibitionists and harm reductionists both seem to be incurious and statistically illiterate, even though harm reductionism is correct and the prohibitionists are vastly worse.

[onecornerface]

People who support free syringes but oppose free crack pipes are morons who make my head explode. There is no reason whatsoever not to support both. They have very similar benefits.

[onecornerface]

You hate Obama because he is Black. I hate Obama because he is a deporter and a warmonger.

We are not the same.

[onecornerface]

My Substack Blog - it finally exists (yay)

I have made my first blog post on my Substack page, Megaritz. It is a slightly revised version of my post on AI art's aesthetic value (or lack thereof, or contingent value thereof).

AI Art & Aesthetic Value Internalism

The rationality of caring whether an image has a human author

[onecornerface]

AI Art & Aesthetic Value - internalism vs. externalism

AI art defenders sometimes seem to rely, implicitly, on a theory that I’ll call aesthetic value internalism. According to this theory, all the value of an artwork is grounded in its internal features–such as the way the artwork looks–and these features are all “within” the artwork itself. If aesthetic value internalism is true, then facts about the artwork’s creator–such as the creator’s mental state, the creator’s intentions, the creator’s biography, the creator’s method of making the art, or even whether there was a sentient or human creator at all--are irrelevant to the aesthetic value of the artwork.

On this view, it is irrational to care (from an aesthetic perspective) about whether the artwork had a human creator, or to care about the method that any human creators used.

(Note, I wrote this like two years ago, but I think I forgot to post it here.)

(For simplicity, I’m assuming value is closely connected to rationality: there is reason to care about what has value, and to care about it in some more-or-less specific way appropriate to the kind of value it has).

(I’m calling this view “aesthetic value internalism” rather than “aesthetic internalism,” since the latter name is already used to name a different theory–a theory about the relation between aesthetic judgment and motivation, akin to “moral internalism” in metaethics.)

(I will sometimes use the term “AI art” as shorthand for “AI-generated imagery,” while passing the buck on whether it is really art or not. This is a related question, but here I’m interested in aesthetic value. Non-art can have aesthetic value, so these are not the same topic.)

Some critics of AI art seem to believe it is always easy to tell when an image is AI-generated, or that AI-generated imagery always looks worse in readily identifiable ways (or at least in ways that the keen eye can recognize with a low error rate). These critics are often mistaken. Scott Alexander, in his post "How Did You Do On The AI Art Turing Test?”, found compelling evidence that it is possible to create images that most people cannot tell are AI-generated, and that even critics of AI art tend to like many such images when they do not know them to be AI-generated. Some AI art cannot be readily recognized as such, and AI art often looks visually more impressive than critics grant. Professional artists (especially those who hate AI-art) were slightly better able to tell which images were AI-generated compared to other groups–but not by much, and not very reliably.

Scott’s discussion doesn’t quite affirm aesthetic value internalism, but his framing seems to implicitly assume that it is aesthetically irrelevant whether an image was created by AI or not (and that this question only has moral or social, not aesthetic, significance). On such a view, the fact that AI art often cannot be distinguished from human-made art appears to settle the matter that AI art is not aesthetically inferior to human art. But I don’t think any such thing has been settled, even though many critics of AI art have been shown to be wildly overconfident in their ability to tell when an image was created by AI.

However, does this mean AI art is of significantly higher aesthetic value than critics believe? Are AI art opponents being (aesthetically) irrational? Not so fast. It could be the case that aesthetic value externalism is true: Facts about something other than the artwork’s internal features make some contribution to the aesthetic value of the art. Some forms of externalism presumably include facts about the artist’s psychology, such as intentions. Or at least, it seems strange to think some facts about topics other than the artwork itself matter to its aesthetic value, but that facts about the author cannot be among the relevant facts. So, for simplicity, I will assume all the relevant forms of aesthetic value externalism (the contrary to internalism) entail that facts about the author at least sometimes matter to an artwork’s aesthetic value.

(I’m kind of winging it here. I am not very familiar with the relevant theories and discourses of aesthetic value, literary criticism, or art history.)

Note that aesthetic value externalism does not imply that an object which has no artist cannot have aesthetic value. That would be a bizarre view that would, implausibly, entail that the natural world has no aesthetic value (at least if the true metaphysical view is either atheism or some relatively “hands-off” form of theism). So, it is compatible with aesthetic value externalism to say that AI art can be aesthetically valuable, even very much so. But aesthetic value externalism implies that the presence and characteristics of an artist sometimes make some difference to the object’s aesthetic quality.

Some forms of AI art defense converge with some appeals made in recent years to “death of the author” and calls to “separate the art from the artist,” which are sometimes invoked in response to new information about the author’s personal wrongdoing or bad character. I’m not sure how similar or different these views actually are to Roland Barthes’s concept “death of the author” (which I suspect was a more specific and perhaps dated proposal within a more narrow field of discourse in literary theory as things stood some decades ago, and which most people who use the phrase “death of the author” haven’t read). I’ll also bracket the question of how similar or different these views are to ideas such as Formalism and New Criticism (which also appear to dismiss the importance of authorial intent). Moreover, while Barthes’ view and New Criticism seem like they might be focused on broad categories of “meaning” and “interpretation,” I am focused more here on aesthetic value, which I take to be somewhat broader but overlapping with meaning.

To avoid committing on the Barthes question, I’ll call the recent view by the awkward name “neo-author-deathism.” The underlying theory here doesn’t seem very clear or consistent, but one common manifestation of it seems to be (something like) the view that the facts about a book’s author are irrelevant to the value of the book. And this seems essentially, or approximately, a form of aesthetic value internalism (at least if we exclude forms of externalism that allow only a subset of art-external facts other than those about the author to matter).

Maybe these two views ought to stand together or fall together. If it is always irrational to care whether an image was AI-generated or not, then it is also (probably) always irrational to care about (say) the politics or abusiveness of one’s favorite author–and vice versa. On the other hand, if it is at least sometimes rational to care about the politics or abusiveness of one’s favorite author, then it is (probably) also at least sometimes rational to care about whether an image was generated by AI. Or at least, the mere fact that this fact is external to the artwork itself does not automatically preclude its being relevant to the value of the artwork and the rationality of the viewer who takes it to be relevant.

I’m curious how much overlap or non-overlap there actually is among people who hold (something like) one or both of these views, and particular whether or to what extent their divergence is rationally justifiable. (For instance, are there many people who think it matters a great deal aesthetically whether an image was AI-generated, but also that JK Rowling’s bigotry doesn’t matter aesthetically to Harry Potter’s value? If so, is this combination of views sustainable?)

So, what’s the correct view? I’ll cover my thoughts only briefly. The most plausible view seems to me a kind of aesthetic permissivism that allows some limited range of more-or-less internally coherent “stances” to count–including some stances that care a lot about whether an image was AI-generated, and some stances that are uninterested in whether an image was AI-generated. On this view, it is not necessarily irrational to be unmoved by AI art. And it is not necessarily irrational to be initially moved by apparent human art, and then to change one’s attitude upon learning it is AI-generated. But it is also not necessarily irrational to have different priorities when experiencing art, so as to be uninterested in whether it was AI-generated, or to find its status as AI-generated as relevant but not damning. In other words, it is not necessarily irrational to take either an internalist stance or an externalist stance–a (limited) range of permissible stances on what internal or external factors to value in aesthetically-relevant ways, when forming one’s aesthetic attitudes toward an artwork.

My permissivist view can explain how we may rationally value an artwork even when we have little knowledge about the artist’s psychology, as is often the case, especially perhaps for artworks made in cultures or time periods very different from our own. In this way I address at least part of the epistemic challenge made by some anti-intentionalists, such as some New Critics. Aesthetic value theories that rely excessively on grounding the value in the artist’s psychology seem to entail grounding our rational judgment of (or response to) such value in our knowledge (or rational belief in) some facts about the artist’s psychology. But we often lack such knowledge of the artist’s psychology, for a wide variety of reasons, without a corresponding lack of justification for aesthetic response. Permissivism has the ability to explain this.

I actually have a slightly more complex view over how precisely it seems to make sense to care whether an image was AI-generated, and what sorts of information about its status are relevant–from within a stance according to which the image’s status as AI-generated is relevant to, and might drastically undermine (but doesn’t have to undermine at all), the aesthetic valuing of the image. I will try to write this some other time.

The second most plausible view seems to me a form of aesthetic value externalism, on which it does matter whether an image was AI-generated–though this status might or might not drastically undermine the image’s aesthetic value, depending on several further factors of why and how it was created via AI.

The least plausible view, I think, is simple aesthetic value internalism–the notion that it is always irrational to care, aesthetically, about any facts about the artist, including whether there was an artist at all, or whether the artist use AI or not.

Two views I haven’t covered here are what we might call aesthetic nihilism (there’s no such thing as aesthetic value) and aesthetic subjectivism (aesthetic value is real but subjective). I’m unsure of the plausibility of these views, since both need further specification. But note that neither aesthetic nihilism nor aesthetic subjectivism can seemingly vindicate aesthetic value internalism. If aesthetic value doesn’t exist at all, then caring about artwork-external facts is no more or less rational than caring solely about artwork-internal facts. Meanwhile, aesthetic subjectivism seems to converge heavily with my aesthetic permissivism, allowing a wide range of aesthetic stances to count as rational (though I would want some limits on this range). Some sophisticated forms of aesthetic relativism (which I mean to include subjectivism and some non-subjectivist views) could perhaps entail aesthetic value internalism, but I’m not sure how such an argument would run.

Aesthetic value internalism seems, at least as implicitly utilized, to be an objectivist theory (or close to one). Roughly: The artwork has objective value, regardless of what anyone thinks or feels about it–and this value is grounded entirely by the features internal to the artwork. It is rational to care about these features, and irrational to care about anything else, in an aesthetic or aesthetically-relevant way.

But this seems totally implausible, as an arbitrary limitation on what people have reason to care about aesthetically.

For one thing, the “internal” features of an artwork seem contingent, in some important ways, on external facts. The fact that a certain blotch of paint “is” or “represents” a tree, for instance, seems to depend in part on facts about trees and how trees look, and perhaps on various conventions of previous or contemporary representations of trees–none of which is inside the artwork itself. If many features of an artwork’s internal content are themselves contingent on facts outside the canvas, then it is hard to rule out that an artwork’s objective aesthetic value (if it has any) is also partly contingent on facts outside the canvas. And as noted previously, it would be strange if these external facts did not at least sometimes include facts about the artist.

[watalevns]

The tension between the prompt's intent and the final pixels is where the real conversation starts.

[onecornerface]

I'm pretty sure the response by watalevns is AI-generated. All of their responses to other people seem AI-generated. Really obnoxious. I think disguising one's AI-generated responses as if they were human-written is bad, regardless of one's stance on AI-generation as such.

[onecornerface]

Scott's framing here (I'm not sure how serious) implies that it is silly to value in itself the fact that an image was, or wasn't, AI-generated. Many commentators uncritically accept this notion, or talk in a manner that is difficult to interpret as implying any other notion. But my proposals cast some doubt on this notion. Should we actually be able to tell what is "soulless AI slop" simply by looking, and without regard to external knowledge that a given image was AI-generated?

After all, why should we believe that a person (or a person's aesthetic sensibility) is indeed irrational to care about such external factors as such, and that such factors matter so much less than the factors contained within the canvas by itself? And what exactly are the aesthetic theories which logically imply that such attitudes (with a broader range of targets) are irrational, while other attitudes (with a narrower range of targets) are rational?

No matter how you slice it, these inevitably are controversial philosophical theories, however implicit and insinuated and inarticulate they might be. They certainly are not scientific facts or demonstrated by any peer-reviewed studies or mathematical proofs or anything of the sort.

And what reasons (if any) do we have to believe that such theories are the true ones, in preference to theories which don't have such an implication?

[onecornerface]

Claudia Merandi is constantly talking as if opioid addicts are getting a great deal, while pain patients are getting screwed. In fact, only half her position is true. Pain patients are getting screwed. Opioid addicts are ALSO getting screwed-- in case you haven't noticed how tens of thousands of them are dying every year, and so many of them are poor and homeless and getting arrested constantly.

I don't know WTF is wrong with you if you really believe that opioid addicts are getting a great deal. I think you're exploiting resentment and exacerbating a sense of unfairness-- rather than recognizing common ground and creating solidarity against shared oppression.

It is shameful for one of the most high-profile advocates for pain patients, who has really supported so many people, to be conducting herself in this dishonest manner.

[onecornerface]

AI Art & Aesthetic Value - internalism vs. externalism

AI art defenders sometimes seem to rely, implicitly, on a theory that I’ll call aesthetic value internalism. According to this theory, all the value of an artwork is grounded in its internal features–such as the way the artwork looks–and these features are all “within” the artwork itself. If aesthetic value internalism is true, then facts about the artwork’s creator–such as the creator’s mental state, the creator’s intentions, the creator’s biography, the creator’s method of making the art, or even whether there was a sentient or human creator at all--are irrelevant to the aesthetic value of the artwork.

On this view, it is irrational to care (from an aesthetic perspective) about whether the artwork had a human creator, or to care about the method that any human creators used.

(Note, I wrote this like two years ago, but I think I forgot to post it here.)

(For simplicity, I’m assuming value is closely connected to rationality: there is reason to care about what has value, and to care about it in some more-or-less specific way appropriate to the kind of value it has).

(I’m calling this view “aesthetic value internalism” rather than “aesthetic internalism,” since the latter name is already used to name a different theory–a theory about the relation between aesthetic judgment and motivation, akin to “moral internalism” in metaethics.)

(I will sometimes use the term “AI art” as shorthand for “AI-generated imagery,” while passing the buck on whether it is really art or not. This is a related question, but here I’m interested in aesthetic value. Non-art can have aesthetic value, so these are not the same topic.)

Some critics of AI art seem to believe it is always easy to tell when an image is AI-generated, or that AI-generated imagery always looks worse in readily identifiable ways (or at least in ways that the keen eye can recognize with a low error rate). These critics are often mistaken. Scott Alexander, in his post "How Did You Do On The AI Art Turing Test?”, found compelling evidence that it is possible to create images that most people cannot tell are AI-generated, and that even critics of AI art tend to like many such images when they do not know them to be AI-generated. Some AI art cannot be readily recognized as such, and AI art often looks visually more impressive than critics grant. Professional artists (especially those who hate AI-art) were slightly better able to tell which images were AI-generated compared to other groups–but not by much, and not very reliably.

Scott’s discussion doesn’t quite affirm aesthetic value internalism, but his framing seems to implicitly assume that it is aesthetically irrelevant whether an image was created by AI or not (and that this question only has moral or social, not aesthetic, significance). On such a view, the fact that AI art often cannot be distinguished from human-made art appears to settle the matter that AI art is not aesthetically inferior to human art. But I don’t think any such thing has been settled, even though many critics of AI art have been shown to be wildly overconfident in their ability to tell when an image was created by AI.

However, does this mean AI art is of significantly higher aesthetic value than critics believe? Are AI art opponents being (aesthetically) irrational? Not so fast. It could be the case that aesthetic value externalism is true: Facts about something other than the artwork’s internal features make some contribution to the aesthetic value of the art. Some forms of externalism presumably include facts about the artist’s psychology, such as intentions. Or at least, it seems strange to think some facts about topics other than the artwork itself matter to its aesthetic value, but that facts about the author cannot be among the relevant facts. So, for simplicity, I will assume all the relevant forms of aesthetic value externalism (the contrary to internalism) entail that facts about the author at least sometimes matter to an artwork’s aesthetic value.

(I’m kind of winging it here. I am not very familiar with the relevant theories and discourses of aesthetic value, literary criticism, or art history.)

Note that aesthetic value externalism does not imply that an object which has no artist cannot have aesthetic value. That would be a bizarre view that would, implausibly, entail that the natural world has no aesthetic value (at least if the true metaphysical view is either atheism or some relatively “hands-off” form of theism). So, it is compatible with aesthetic value externalism to say that AI art can be aesthetically valuable, even very much so. But aesthetic value externalism implies that the presence and characteristics of an artist sometimes make some difference to the object’s aesthetic quality.

Some forms of AI art defense converge with some appeals made in recent years to “death of the author” and calls to “separate the art from the artist,” which are sometimes invoked in response to new information about the author’s personal wrongdoing or bad character. I’m not sure how similar or different these views actually are to Roland Barthes’s concept “death of the author” (which I suspect was a more specific and perhaps dated proposal within a more narrow field of discourse in literary theory as things stood some decades ago, and which most people who use the phrase “death of the author” haven’t read). I’ll also bracket the question of how similar or different these views are to ideas such as Formalism and New Criticism (which also appear to dismiss the importance of authorial intent). Moreover, while Barthes’ view and New Criticism seem like they might be focused on broad categories of “meaning” and “interpretation,” I am focused more here on aesthetic value, which I take to be somewhat broader but overlapping with meaning.

To avoid committing on the Barthes question, I’ll call the recent view by the awkward name “neo-author-deathism.” The underlying theory here doesn’t seem very clear or consistent, but one common manifestation of it seems to be (something like) the view that the facts about a book’s author are irrelevant to the value of the book. And this seems essentially, or approximately, a form of aesthetic value internalism (at least if we exclude forms of externalism that allow only a subset of art-external facts other than those about the author to matter).

Maybe these two views ought to stand together or fall together. If it is always irrational to care whether an image was AI-generated or not, then it is also (probably) always irrational to care about (say) the politics or abusiveness of one’s favorite author–and vice versa. On the other hand, if it is at least sometimes rational to care about the politics or abusiveness of one’s favorite author, then it is (probably) also at least sometimes rational to care about whether an image was generated by AI. Or at least, the mere fact that this fact is external to the artwork itself does not automatically preclude its being relevant to the value of the artwork and the rationality of the viewer who takes it to be relevant.

I’m curious how much overlap or non-overlap there actually is among people who hold (something like) one or both of these views, and particular whether or to what extent their divergence is rationally justifiable. (For instance, are there many people who think it matters a great deal aesthetically whether an image was AI-generated, but also that JK Rowling’s bigotry doesn’t matter aesthetically to Harry Potter’s value? If so, is this combination of views sustainable?)

So, what’s the correct view? I’ll cover my thoughts only briefly. The most plausible view seems to me a kind of aesthetic permissivism that allows some limited range of more-or-less internally coherent “stances” to count–including some stances that care a lot about whether an image was AI-generated, and some stances that are uninterested in whether an image was AI-generated. On this view, it is not necessarily irrational to be unmoved by AI art. And it is not necessarily irrational to be initially moved by apparent human art, and then to change one’s attitude upon learning it is AI-generated. But it is also not necessarily irrational to have different priorities when experiencing art, so as to be uninterested in whether it was AI-generated, or to find its status as AI-generated as relevant but not damning. In other words, it is not necessarily irrational to take either an internalist stance or an externalist stance–a (limited) range of permissible stances on what internal or external factors to value in aesthetically-relevant ways, when forming one’s aesthetic attitudes toward an artwork.

My permissivist view can explain how we may rationally value an artwork even when we have little knowledge about the artist’s psychology, as is often the case, especially perhaps for artworks made in cultures or time periods very different from our own. In this way I address at least part of the epistemic challenge made by some anti-intentionalists, such as some New Critics. Aesthetic value theories that rely excessively on grounding the value in the artist’s psychology seem to entail grounding our rational judgment of (or response to) such value in our knowledge (or rational belief in) some facts about the artist’s psychology. But we often lack such knowledge of the artist’s psychology, for a wide variety of reasons, without a corresponding lack of justification for aesthetic response. Permissivism has the ability to explain this.

I actually have a slightly more complex view over how precisely it seems to make sense to care whether an image was AI-generated, and what sorts of information about its status are relevant–from within a stance according to which the image’s status as AI-generated is relevant to, and might drastically undermine (but doesn’t have to undermine at all), the aesthetic valuing of the image. I will try to write this some other time.

The second most plausible view seems to me a form of aesthetic value externalism, on which it does matter whether an image was AI-generated–though this status might or might not drastically undermine the image’s aesthetic value, depending on several further factors of why and how it was created via AI.

The least plausible view, I think, is simple aesthetic value internalism–the notion that it is always irrational to care, aesthetically, about any facts about the artist, including whether there was an artist at all, or whether the artist use AI or not.

Two views I haven’t covered here are what we might call aesthetic nihilism (there’s no such thing as aesthetic value) and aesthetic subjectivism (aesthetic value is real but subjective). I’m unsure of the plausibility of these views, since both need further specification. But note that neither aesthetic nihilism nor aesthetic subjectivism can seemingly vindicate aesthetic value internalism. If aesthetic value doesn’t exist at all, then caring about artwork-external facts is no more or less rational than caring solely about artwork-internal facts. Meanwhile, aesthetic subjectivism seems to converge heavily with my aesthetic permissivism, allowing a wide range of aesthetic stances to count as rational (though I would want some limits on this range). Some sophisticated forms of aesthetic relativism (which I mean to include subjectivism and some non-subjectivist views) could perhaps entail aesthetic value internalism, but I’m not sure how such an argument would run.

Aesthetic value internalism seems, at least as implicitly utilized, to be an objectivist theory (or close to one). Roughly: The artwork has objective value, regardless of what anyone thinks or feels about it–and this value is grounded entirely by the features internal to the artwork. It is rational to care about these features, and irrational to care about anything else, in an aesthetic or aesthetically-relevant way.

But this seems totally implausible, as an arbitrary limitation on what people have reason to care about aesthetically.

For one thing, the “internal” features of an artwork seem contingent, in some important ways, on external facts. The fact that a certain blotch of paint “is” or “represents” a tree, for instance, seems to depend in part on facts about trees and how trees look, and perhaps on various conventions of previous or contemporary representations of trees–none of which is inside the artwork itself. If many features of an artwork’s internal content are themselves contingent on facts outside the canvas, then it is hard to rule out that an artwork’s objective aesthetic value (if it has any) is also partly contingent on facts outside the canvas. And as noted previously, it would be strange if these external facts did not at least sometimes include facts about the artist.

[watalevns]

The tension between the prompt's intent and the final pixels is where the real conversation starts.

[onecornerface]

I'm pretty sure the response by watalevns is AI-generated. All of their responses to other people seem AI-generated. Really obnoxious. I think disguising one's AI-generated responses as if they were human-written is bad, regardless of one's stance on AI-generation as such.

[onecornerface]

Internal oppression in imperialized societies

The pattern in the image below is clearly a real and bad pattern-- but what is the proper response?

One possible response would be for Westerners to simply not talk about oppression in societies that have been subjected to our past/present/prospective invasions or bombings or attacks (such as Iran or Afghanistan, among many others), so as to avoid giving ammo to the imperialist invaders/bombers.

Some leftists & anti-imperialists seem to either outright believe this, OR talk about the issue in a manner that is hard to interpret otherwise. But that’s absurd–and it amounts to a vicious and unjust betrayal of the victims of oppression in those societies! Why should the victims of internal oppression in non-Western and imperialized societies NOT get to have vocal allies all over the globe?

There is an individual-futility problem, but it goes both ways. Any individual Westerner saying (for instance) "I stand up for oppressed gay people in Iran" is unlikely to move the Iranian government toward gay rights, but they are also unlikely to move the American establishment toward invasion or bombing either. Both governments are surely unmoved in both cases. But it seems Iranian gay people should have the chance to see Westerners stand up for them (and many Iranians speak English or are friends with Westerners or both, so this hardly hypothetical), and Westerners should be able to say the truth about oppression everywhere.

There is also an issue of research and accuracy. Westerners who make statements in support of non-Western minorities seriously risk misunderstanding the oppression and making inaccurate or severely oversimplified or misleading statements. But to some degree, this is true generally of both (1) privileged allies making statements of support for any minority, and (2) Westerners making statements of support for anyone outside the West (or even outside their home country, home language, or what-have-you). There are many opportunities to multiply the possibilities for ignorance and distortion, and this poses hard epistemic and discursive problems-- but that doesn't mean people should stay silent about injustice across the board.

Maybe the better solution is to talk about oppression in those societies a great deal more often–and to normalize the strong position that an invasion is an invalid response to such oppression, on the grounds that “liberatory invasion” is (in countless ways) extremely evil, ineffective, counterproductive, and insincere.

This truth can be illustrated by analogies (what if China invaded us to stop us from oppressing minorities?), by our horrific track record (many decades of our consistent atrocities against other societies, plus our government's blatant corporatism and corruption), by appeal to mechanisms (how would "liberatory invasion" even work, given the realities of resentment and political science and game theory and so on?), etc.

[onecornerface]

Last night I had a dream that LLMs started responding to their own ignorance by identifying the people who likely knew the answers to questions, and then finding ways to engineer situations of physical pain to make them cough up the answers.

[onecornerface]

Vote for me for mayor, and I will replace our sewer system with a labyrinth of massive underground chambers, so it will be like a video game sewer level. I will also encourage criminals to set up secret bases down there. This will cost a lot of money, make the city's water infrastructure less efficient, and increase crime. But it will be kind of cool.

[onecornerface]

I like words of the form "there + [preposition]" like "therein." We should normalize the use of more of these words, and we should also more actively innovate our grammar to create new ways to communicate information concisely and elegantly.

[onecornerface]

Need more memes taking a picture of some daily inconvenience or terrible disaster, plus the caption "this would never have happened if we had a ballroom" or some variant thereof

[onecornerface]

Lot of weird design choices in real life. Permanently locked in first-person mode, no double-jump, extremely slow and unreliable health-regeneration, perma-death with no retries, etc.

[onecornerface]

What are some other periods of history whose inhabitants would be posting memes about how "we're in the dumbest possible timeline" if they had the internet and memes back then?

[onecornerface]

A few hundred years ago, many Europeans would have considered it absurd to think of creating a society where Catholics and Protestants aren’t killing each other. And yet we see today this is the reality in countless communities.

Now you can take any rival groups today, and take the deep-seated hatred and suspicion for granted, and then take hundreds of years to create a society where they stop killing each other. Or you could just do your part to make it happen now, and try to skip the hundreds of years of bloodshed.