@07mk's banner p

07mk


				

				

				
2 followers   follows 0 users  
joined 2022 September 06 15:35:57 UTC

				

User ID: 868

07mk


				
				
				

				
2 followers   follows 0 users   joined 2022 September 06 15:35:57 UTC

					

No bio...


					

User ID: 868

This is the same way that an art student "copies" the art of other artists by the light reflecting off the picture s activating their rods and cones which get converted to signals in the neurons in their brains which produce the perception of the image in their consciousness. By your definition of copying, all perception of media is copying; e.g. every time I listen to Bohemian Rhapsody, I'm copying the work of Queen. Which is a fine definition, but is also not exactly a common one and one that loses a lot of the meaning that's intended when people use that word.

So much of the hatred directed at AI art I see just makes me think of this old Futurama joke.

Relevant lines:

"With all that Salmonella and me been through, her sound is unique. All your fancy technology will NEVER be able to copy this guitar."

"Using my fancy technology, I can make an exact copy of this guitar."

Of course, duplicating a physical object is a different endeavor, but the fundamental phenomenon is the same.

This post seems like it's culture war waging and doesn't belong here. However, I wanted to say that I love this passage as a beautiful description of what practicing to develop drawing skills (or really, any kind of physical skills) really is, on a fundamental level.

For a mangaka, «fish-eye» means years of wrestling with built-in visual cognitive biases, painstaking fiddling in awkward 2D, selecting, combining and projecting primitives, generating and tracking pencil trajectories, iterating a loop of deliberately memorized algorithms with elements of pseudorandom – assembling a cognitive chimera from the elements of thought amenable to control.

Shouldn't this apply to any other part of the industry as well? Anyone who's paid to produce artwork of this sort, whether they be a comic book artist or TV show character designer or video game advertiser or etc. could make use of AI art software to help at various steps along the way going from a blank canvas to the final product. To me, AI art software seems akin to Photoshop or a drawing tablet in how it would allow existing producers of 2D static art to be more efficient or effective in their work.

Thank you for this write-up. Things seem to be happening very quickly with a lot of volatility, so it's hard to say where things will go, but reading this makes me a little more pessimistic about the future of AI art. I had hoped that Stable Diffusion would offer a sort of "AI of last resort" that users could rely on that was largely free from authoritarian control, but how AUTOMATIC1111 was treated, and the comment about "handling edge cases" are both very concerning. I'm reminded of how dreams of a democratic open web from the 90s and early 2000s have been dashed in the intervening decades through most communication being controlled by a handful of censorious companies, with the logistics of setting up alternatives being too high of a hurdle to overcome for actual open alternatives to exist in a meaningful manner. That said, I don't know enough about AI art software to tell if a similar thing is likely to happen here; at the least, the logistics of collecting and collating the data required to train such software seems like a major hurdle, even before getting into the legal issues.

Also:

And as an extra bonus, the coomers using SD have gone into high alert after Emad remarked in an interview that the release of SD V1.5 would be delayed to "handle edge cases" & (de)bias the model against "topless women".

That link goes to https://todo.later/. Were you intending to replace that with a different URL?

I see, it seems that I misunderstood you, and your point seems a good one. Could it be that they're principled but lazy? Which, to be fair, is really just another way of saying nihilistic.

That sounds wrong and likely delusional and stupid, but not nihilistic, at least not without extra steps. According to SocJus ideology, naming something like Coq would cause some people, especially women and people of color, to feel uncomfortable and thus less encouraged to join projects that might use it. Under this framework, changing the name is a way to make the project meaningfully more inclusive.

Also:

It is just simple wording or including a CoC(note the acronym for Code of Conduct) . It is not consistent or helping the project in a meaningful way to be more inclusive.

I have to wonder how long it will be before Codes of Conduct will have to be renamed for forming an acronym that is similarly harmful to inclusivity according to SocJus ideology.

That's a great point and in retrospect so obvious that I'm surprised it didn't occur to me before.

This sort of ties back into the original topic about principles. One thing that has seemed obvious to me for a while is that, given everything we know about psychology and the science of self-serving biases, if someone actually wants to be principled, they must meaningfully question their own principles with the presumption that the reason they adopted those principles is because doing so is beneficial to themselves, not because those principles are actually the good/right/just principles. Without doing so, they're merely someone who has naked self-interest as their principle and are really good at hiding that from themselves.

Given that, someone who really cares about being principled would fight for their principles in cases that offer them little benefit with just as much (actually, even more, given what we know about biases in perception) vigor as in cases that offer them a lot of benefit, and likewise in cases where following those principles would harm them. It's only by taking a costly action by focusing on things that would be of little personal benefit at the cost of things that would be of great personal benefit or by outright calling for something that would harm oneself that one can actually prove to oneself that one believes in those principles rather than adopting them out of convenience.

Having observed the culture wars over the past decade or so, I must admit that my estimation for how principled people tend to be has dropped considerably.

As an aside: a couple of the comments here about the fact that "actor" is a job where belonging to a certain race (or more broadly just how one looks) is sometimes itself a qualification reminded me of a Freakonomics podcast episode I'd heard probably 5+ years ago which mentioned that "restaurant server" is also one of those jobs. Specifically for restaurants of ethnic cuisine, where the servers are expected to look like people from the region where the food is from.

Before I'd ever heard the episode, I had certainly noticed that Chinese/Korean/Japanese restaurants I'd gone to tended to have servers who were Asian. I'd always chalked it up to the restaurant owners often being from the food's region and hiring family and friends as servers, but I had also noticed more than a few times that servers' actual heritage didn't match up (e.g. Vietnamese server at a Korean restaurant - something many/most American customers wouldn't notice without purposely looking for it). From what I remember from the podcast episode, it just seemed like an industry-wide thing, that the server vaguely looking to a typical customer like they come from the food's region was an important factor when hiring them. It seems to be one of those carve-outs to the principle of non-racial-discrimination-in-hiring-practices that gets talked about even less than acting.

I think adherence to principles that cause one to do good is good and consistency to principles that cause one to do evil is evil. I would rather someone with evil principles be inconsistently evil (and thereby inconsistently good) than be consistently evil.

Frankly I find this perspective... well, not bizarre, but pure hubris. It has an underlying assumption that one can reliably distinguish between principles (or more specifically actions) that are good versus ones that are evil. Which is all well and good when dealing with cartoons and their cartoon villains and heroes, but has rather minimal applicability in the real world. However, I do understand that believing that one's own principles are the good ones is a very common - possibly universal - bias, which is why I don't find it bizarre.

Furthermore, leaving aside moral judgments, it ignores the logistical aspects. If I observe someone else consistently following principles that I consider to be evil, then I can conclude that they are following principles that they consider to be good, and someone who follows principles they consider to be good is someone whose behavior I can predict by learning their principles. Someone who follows principles that I consider to be evil but does so randomly is someone whose behavior I can't reliably predict (assuming there isn't some other principle that they're actually following I can learn about), which makes interacting with them intrinsically unpredictable. Someone who consistently follows principles that I consider evil are people I can learn from and live with, but someone who inconsistently follows principles that I consider evil are wild cards.

I'll admit that I'm not an academic in the field, but I have done the research within the role of activist and found that anti-racism isn't an entire academic field - it's an entire pseudo-academic field, and I also found the prominence of DEI has basically no root in someone looking into this and finding that it's actually the case (aside: it was largely because I did this research that I'm no longer in that role of activist). It's only because I'm not an academic in the field that I leave open the possibility that I missed something despite my having done the research to the best of my abilities already, which is why I ask if the research exists.

And the point you keep seeming to ignore is the context of fictional characters on screen. The leap from "interpersonal attraction to people similar to oneself" to "can't identify with [fictional depictions of] people who have other skin colors than themselves" is one that needs actual empirical support, not just "fairly obvious point that is a very strong corollary." That's simply not how science works, especially in the realm of something "soft" like sociology/psychology.

I never had any problems identifying with Goku just because he was Japanese.

This is an interesting case, since Goku is actually an alien, and the Earth of Dragon Ball is far removed from the real Earth, with none of the same countries as real Earth or even countries that are equivalent to real countries on real Earth. Goku's backstory was based partly off of Journey to the West, which is an old Chinese story, while his character design was by a Japanese artist following Japanese manga standards, giving him a highly stylized appearance that evokes a Japanese person to the audience. Goku's backstory was also based partly off of Superman, who's another alien who takes the appearance of a human from a particular country (USA in this case).

I don't think that link shows that we gravitate towards people who are similar to us in terms of race, particularly in the realm of watching fictional characters on screen. You can't make the leap from the broad tendency of interpersonal attraction towards similar people (from the meta-study that was the 1st link in that Psychology Today post) to there, not without going through many steps that each require empirical support.

I think Electro commented in that film that there must be a black Spiderman somewhere out there in the multiverse, in a shout out to Miles Morales. Though that's obviously not a direct cameo.

We gravitate to people that are like us [in terms of race]. Also, kids are dumb. They are going to internalize and notice when no characters look like them [in terms of race].

This is the kind of thing that's easy to claim and also easy to believe, but I've yet to see it actually supported through evidence. Given that, I don't see why anyone should take this claim any more seriously than any other unsupported claim.

To the point of advocates, I was cajoled into seeing the recent spiderman movie and I remember there was a cameo of some black superhero, and all of the black kids in the audience went nuts over it. And it was clear in that moment that there's a compelling need, to some extent, for more representation of x demographic, because, for instance, it can't be positive to grow up watching superhero movies and none of them look like you.

(Emphasis added). I've seen this line of thinking used to justify race/gender-swapping a lot, and I'm skeptical that it is the case in general. Certainly, for some people some of the time, it's true (much like pretty much anything), but I've yet to see anyone do the hard work of checking this through scientific research. My suspicion is that this line of thinking is supported by means of projection by people who do feel like other human beings of different skin color/gender/etc. are less empathetic, but ultimately sans empirical evidence, I'm agnostic. It also seems to me that, to whatever extent people suffer due to, say, superheroes not being of the same skin color/gender/etc. as themselves, it's likely to be socially modulated, akin to how white Americans in the 1920s might have suffered from sharing a school with black people, but white people in the 2020s don't.

If Kanye started spouting rants about how black people should kill all whites, I suspect the same people outraged now would say "You have to understand that he is mentally ill and his words shouldn't be taken at face value; I hope he gets the help that he needs."

I don't think this is quite right; mental illness wouldn't have to be invoked as an excuse/justification at all. The justification would simply be that this is the righteous anger of a black man born in a White Supremacist society that has oppressed him throughout his life. Sure, he may appear a little overenthusiastic in his quest for justice, but it is not our place to judge the way an oppressed person reacts against his oppressors.

At least, that's my belief based on how I've seen chants like "kill all men," "kill all whites," "kill all cis" supported throughout the years.

I've never once done squats with a partner. A squat rack is really all you'd need, which most gyms have. I'd definitely recommend trying them out and incorporating them into your workouts, since they're a very useful compound lift that works on a lot of your muscle groups. The deadlift is another one that would be useful if you're not already doing it, for similar reasons. And a deadlift is safe without any equipment or partner. Just gotta get the form right to not hurt the back.

This is why I wrote that the results would be identical on video, not that the results would be identical as a film or as a piece of art.

The slapping man and the Front Man are both major movie stars in Korea (the Front Man has even crossed over to major Hollywood productions) essentially making cameo appearances in the show, so it's not surprising so many people found the former so attractive. I'll admit I found the protagonist fairly homely, especially by leading man standards, so I'm surprised to read a couple people say he was attractive.

OK, I had to think about this for a while, but I think I follow. You're really only looking at those who are so far away from passing that they would fail to pass 100 times out of 100, and looking at the distribution of how badly they fail at passing? I think that makes sense, and you're probably correct in your inference.

I’m going to compare this, for no reason at all, to Squid Game. Attractive, interesting actors with strong expressions who we emotionally relate to.

Did Squid Game really have that many attractive actors? I thought the one North Korean defector was very attractive (IIRC the actor is a model), but I found the others pretty average, possibly even below average by major-production-actor standards. There's the undercover cop too, I suppose, though he's barely a major character.

As an aside, Tom Cruise might be weird as fuck, but that mother fucker knows how to make a fucking movie. Sure, he's no Stanly Kubrick. But in a world of complete and utter drek, someone autistically repeating the steps movies used to take to be good, whether he understands what he's doing or not, comes off like a savant.

I legit think Cruise is one of the best filmmakers of all time, at least within the role of "stunt actor." How many people have ever strapped themselves to the outside of a plane that was taking off, much less done it 8 times just for the sake of creating entertainment? Given the continual development of CGI technology, he also might be one of the last such figures, since it won't be too long before we can get literally identical results on video without putting real people at such high risk.

What made them think hiring subpar writers, rewriting lore, rewriting characters of one of the most popular fantasy IPs while simultaneously drafting off of the brand was a good idea? It feels like the motive isn't even to make money but solely to push an agenda, but who would do that?

I don't think they thought they were hiring subpar writers. I think they just have a different measure of what a quality writer is compared to the general public. When people who push this sort of agenda say that having representation of people of certain races/genders/sexualities makes a piece of fiction better, I think they actually believe it. They really do believe that if you take an existing successful franchise and tack on a story that appeals to such sensibilities with diverse characters or allegorical plot threads, it makes the work, in some real meaningful way, better. And thus hopefully more successful. And so they prioritize doing just that.

But, of course, there's no such thing as prioritizing everything at once. If you prioritize the messaging, you necessarily put less priority on the actual quality of the work in terms of things like world building, character development, interesting plot, etc. And making a good TV show isn't easy; you can't just de-prioritize those things and expect to end up with good results. But, again, the people making this stuff genuinely believe that the messaging is what makes the TV show good. There's much to be skeptical about when it comes to the narrative of Hollywood/filmmaking leftists being too far in their own echo chambers to understand what appeals to the general public, but the more I observe the leftist echo chambers in my own environments, the more I can believe in such a narrative.

Of course, such delusional beliefs do have to encounter the stark reality of revenue and watch numbers, but culture like this tends to turn slowly, and there's enough money to keep them afloat and going. And the always-dependable narrative that "we didn't fail the audience; the bigoted audience failed us by refusing to watch our show" (examples abound, but the recent Billy Eichner movie comes to mind, as well as the Charlie's Angels reboot-reboot from a couple years back), followed by "but next time, as the march of progress continues on and our side gets more and more vindicated as being the right side of history, the audience will be receptive to our correct notion of what constitutes quality."

was rooted in a misunderstanding about what "hard-R" referred to

Perhaps you were confusing it for the "R-word" or "R-slur," aka "retard/retarded?" I've noticed that there's a wide range in how taboo this slur is even in leftist circles. Where I grew up, the "R-word" would be only 1 level below the "hard-R" in terms of offensiveness, but I've noticed that in other environments with similar levels of leftism, it's used pretty freely without anyone noticing. E.g. I was shocked when I first noticed how much people used it on Twitch of all places, given how strict Twitch tends to be with authoritarian-left language policing.