07mk
1hr ago
I think people generally do think photography is different to painting (less artful even). Obviously there's a lot of creative decision-making involved in a given photo, but not as much as for a given painting.
I'm not sure what "less artful" means, but certainly no one would claim that ai generated images aren't different from paintings, even when they explicitly use painting styles. Much like how photorealistic photographs are different from photorealistic paintings which are different from photorealistic collages which are different from photorealistic CGI renderings, in non-trivial ways. Same would go for any AI generated images.
Whether it's "as much," I'm not sure how it's possible to quantify the amount of creative decisionmaking in a way that can be meaningfully compared like that.
Across a photographer's oeuvre, you start to see more and more evidence of intentionality, and it takes collections and curation to establish your bona fides as a photographer to a greater extent than as a painter.
This phenomenon is quite evident to exist in people who use AI generated images from following anyone who has posted AI generated images for a long time as well. It applies just as well when you take the AI generated part out of it; even Twitter accounts that merely share pre-existing images of any provenance inevitably establish a pattern of intentionality in terms of the images they deem worthy to share, ie curate. I don't know if a curator is an "artist" who deserves "credit" for their "art," but certainly a curator is someone who makes creative decisions.
What I'm saying is that density of micro decisions is a relevant criteria for assigning credit. I'm not suggesting that an ai prompter deserves no credit.
Again, this seems perfectly cromulent to me, but also, I really don't think "deserving credit" is a particularly meaningful thing. People subjectively credit various things for their works, like God or their family, or only their hard work and effort, or their 5th grade teacher, or the barista whose off-hand comment triggered something in their brains, etc. and I don't really have an opinion on that, other than that it doesn't seem worth having an opinion on. My point is just that "creative decisionmaking" isn't a line that cuts between AI generated images and paintings/sketches/etc. and this applies for any other analogous media.
07mk
1hr ago
In this case I think the sense that no skill or taste has been exercised is important. Drawing a picture seems to require some level of effort or skill, which a person has acquired over years of practice. There may be an incoherent feeling that 'resistance' is important to art.
I'm not sure if that's incoherent, but it seems perfectly cromulent to believe that "art" requires some minimum effort and skill, and that base ai generations don't reach that bar, and likewise something like 4'33", which took neither to write and takes neither to perform. But that's a very different and completely orthogonal point from the notion that AI generated media doesn't reflect human creative decisions.
To be fair to Mixtape, I understand that the protagonist is a huge music nerd, and the game is about her last day in her hometown before driving across the country to give her mixtape to some music producer to try to get a job. As someone who had even less interest in music in the 90s than I do now, I don't know how much that would make her knowledge of obscure and 80s bands more justifiable, though. The point that no teenager in the 90s had nostalgia for the 80s strikes me as very very true, though.
With respect to the rewinding a tape via pencil scene, one important aspect of it is that she specifically spins the wrong hole to rewind the tape. One doesn't need to have first-hand experience to know this: basic understanding of physics should tell anyone that you have to spin the reel that has less tape around it, and spin it such it "pulls" the exposed part of the tape towards that side. In the game, she sticks the pencil into the one that has more tape, somehow "pushing" the tape into the other reel instead of out of the cassette, as would actually happen. Perhaps it's a bug that will be patched, but it's fascinating to me that this obvious physical error got through. Reminds me of the Wonder Woman film where she blatantly doesn't protect her feet while rushing some foxholes during WW1.
07mk
14hr ago
I'm not sure what "created art by themselves" would mean specifically, but someone putting in a penny into that robot would certainly have made a creative decision by deciding to use that particular machine or to use any machine at all to create a drawing. He could have inserted the penny into a different machine, or taken that penny and scratched markings onto a piece of paper to create some "art," or, at possibly the most trivial case, he could have also just decided to frame that penny and present it as "art" in itself, and all of those would have involved some level of creative decisionmaking. In each case, the way the final result reflects his decisionmaking would be somewhat different. I.e. if he actually etched onto a piece of paper using the penny, he would have made decisions on where and how hard each marking was, whereas if he just framed the penny, he would have decided which penny to present, but he didn't make any decision on the angles of the curves that form Lincoln's portrait on the penny.
I find myself wanting to say a similar sort of thing about AI, only a lot of people's decisions were involved in training it rather than just one.
Certainly, I'd agree that many people's decisions went into any diffusion model-based AI-generated image, not just that of the person who typed in the prompt. Much like many people's decisions went into a photograph, such as that of the lens manufacturer and the city engineer who deemed that some building's awnings had to be a certain way and the businessman who spit his gum out on the sidewalk one day which turned into a black mark and the kid who decided to throw rocks at a flock of birds seconds before the photographer pressed the shutter.
07mk
16hr ago
The cost would really only be relevant if there was some reasonable expectation for honesty or accuracy on the internet - especially on social media - like there is in a paid transaction, though. And there is no such expectation. In fact, it's such a cliche that the internet is filled with lies that the one thing Abraham Lincoln is loved for, the very reason he was iconic enough to be put on the penny, is that he warned us not to believe everything you read on the internet over a century before the internet even existed.
My first and immediate question upon learning of the whole Mixtape kerfuffle - how does a small indie company get funds to license 20-something actual songs from I imagine what are pretty fucking big labels?
I've heard people hypothesize that this was a pet project of Megan Ellison, who is founder/owner of Annpurna, the publisher of Mixtape (developed by a different Australian company). She is also the daughter of Larry Ellison (a hectobillionaire). I did some basic research, and it seems true, according to what I would consider credible sources, that Megan Ellison owns Annapurna, Annapurna did publish Mixtape, and Megan Ellison did inherit a large trust from her hectobillionaire dad.
Megan Ellison turned 40 this year so she's an elder millennial, though I'd guess that her upbringing wasn't what one would consider typical of a California millennial in the 90s. The narrative I've seen hypothesized is that she wanted a power fantasy where she could live out the stereotypical cliche teenage rebellion that she never could due to her immense wealth and privilege, but that's very much speculative.
EDIT:
As we continue our mission for booze on the way to Camille Cole's party, I require something that fills every possible spectrum of sound, causing the brain to super-aneurysm and explode and die.
This is Love by The Smashing Pumpkins.
I'm a musical philistine, so I didn't quite understand why this was a whopper. Having listened to the song on YouTube just now, it does seem rather repetitive and hitting the same sort of "sound" for the entirety of the song compared to something like, say, Bohemian Rhapsody. Is that why you'd consider it a whopper, possibly generated by an LLM that has no actual understanding of the song and was just putting words together that seem fitting?
07mk
19hr ago
Humans can take credit for AI art to the extent I can claim to be an artist when I pay a human artist to "make me a picture of grapes and some flowers and a skull". If I specify the picture extremely specifically then I start to be able to claim a fraction of the authorship, but it takes a lot for us not to intuit that the dude holding the paintbrush deserves most of the credit. Most ai artists are more like commissioners.
I don't particularly care about "credit," or what "art" means, but the part of this analogy that I find wrong is that a generative AI tool lacks agency like a human, and I don't see a way to bridge the gap until we've got scifi-level AI. A diffusion model or an LLM is "making decisions" in the same way that a bristle of a paintbrush "makes decisions" on where to place the paint on the canvas, i.e. by following the laws of physics, compelled by the human that's actually controlling the tool (whether through typing in a prompt or waving the paintbrush).
07mk
23hr ago
I disagree, it applies just as well to typing the prompt. Why wouldn't it be? Both are processes that aren't really legible if you watch them without witnessing the result, with the result as its output.
Because the text prompt isn't the actual result, unlike in the case of a poem. It's merely a conduit by which the final result is created, transferring the human's creative decisions to the final artifact. Most accurately, the waving of the hand would be analogous to the finger movements AND the prompt, as ephemeral things that exist momentarily in order to transfer the human's intents to the final piece.
07mk
23hr ago
And waving the brush is where the creative decisionmaking is in painting, which is why it's analogous to the prompt (waving the brush would be analogous to the movements for the fingers when we're talking about typing out a poem or essay or somesuch). The painting is merely an artifact that remains as the result of that creative decisionmaking. And that artifact is what people generally consider the actual creative work or "art," not the decisionmaking itself.
07mk
23hr ago
Did they use that song in South Park, like the montage song? I thought that was written for Lease, the Rent parody in the Team America film.
07mk
1d ago
To latch on to this with another recent minor CW-related kerfuffle, I've seen some speculate that this segment from the climax of recently-released video game Mixtape (note that the game footage is real and unedited; the only addition is the guy reacting to it on the left side) was based off an AI-generated script. For me, the negative parallelism "this ain't a catastrophe, it's a warning" feels AI, but I worry that it's become such a commonly known AI "tell" that it's just a false positive. But also, the phrase "this will only be the beginnings of my wicked ways" just feels AI-ish to me in a way that I can't put my finger on. Something about how elaborate and almost forced-poetic it feels, particularly given that it's a teenager in a highly emotionally charged moment. But that could just be bad writing.
Season 5 of Stranger Things was also speculated as having been partially written using AI (apparently the making-of documentary included a shot of the writers having ChatGPT as one of their browser tabs, but the speculation started before, AFAICT), with people making compilations like this one, but beyond the negative parallelisms - which could be false positives - I'm not sure how the other parts feel AI. Again, there's a bit of forced-poetic feel in a lot of it, especially given, again, these are mostly teenagers. But also, awful writing of teenagers that seem way too sophisticated or mature has plagued this series since at least season 4, probably at least season 3.
Does anyone better with words than me have any opinions on this?
07mk
1d ago
Real art is made by an artist, and involves creative decisions. Algorithms can't do that.
All this seems perfectly cromulent, but this doesn't seem to address the difference in AI generated imagery versus, say, a painting. Whether the algorithm is a diffusion model, an LLM, or the physics of molecules of paint, brushes, canvas, etc., the algorithm itself can't make creative decisions. But humans can and do make creative decisions in terms of how to direct those above algorithms to generate images.
I've seen this exact line of thinking brought up many times in discussion about AI art, and I'm confused why people seem to think that that means (modern) AI-generated images (and songs and poems, etc.) aren't the results of human creative decisions. Even putting in a blank or a randomly generated prompt into the first diffusion model one encounters is a creative decision. Even if we take away the AI and posit that the images were just sitting on the floor, poofed into existence by God or aliens or random chance of nature, the decision to share it with others is a creative decision. Until we get to truly agentic AI, any media that's shared is necessarily the result of a human making a creative decision somewhere.
07mk
1d ago
Why wouldn't the result of the AI generation also be the human creation? Saying that the prompt is the "real human creation" is analogous to claiming that "waving a brush in front of you is the real human creation" in a painting, the painting is just the artifact that results from the paint, brush, and canvas atoms interacting with each other as a consequence of that waving. It'd certainly be cool if we got to see that with every painting, though.
07mk
1d ago
The steelman to this argument which, as an atheist, I find the closest to being actually compelling, is that this is not evidence that one's own religion is true, but it is evidence that the correct thing to do is to believe as if one's own religion is true. As someone who wants to be someone who believes things based on how true they are rather than how correct it is to believe something, I can't make the leap, but I can sorta see the sense in such an argument.
07mk
1d ago
I got off the train before new atheism turned into wokeism, thankfully, but I do think it's hard to justify why, other than that my spidey-sense was tingling.
I had a similar experience, and it was always obvious to me why, and I'm actually surprised to read that someone who followed a similar trajectory would be confused as to why for themselves. It's because new atheism was built on things like evidence, science, rationality, and such, but the wokeism stuff were pretty plainly and openly anti-evidence, -science, -rationality, and -such, in favor of personal revelation (i.e. "lived experience") and just faith. New atheism prioritized using logic and reasoning to convince, while wokeism was, again, plainly and openly, against logical argumentation and for bullying and shaming in order to get people to believe/do things.
07mk
1d ago
I heard that Nelson Mandela originally made the girl have braces, but after he died in prison in the 80s, the Berenstein Bears conspired to destroy all copies of the original cut and replace it with one where the woman doesn't have braces.
07mk
1d ago
Reverse-parsimonially, I think the universe would necessarily be more complex with God or god or gods than without, and I think one could make the case that more complex means less boring. Even if you took the most interesting man in the world, he'd be more boring than his alternate universe version that was identical to him in every way except he did one other thing, no matter what that other thing was.
07mk
1d ago
If you take two pixel-by-pixel identical artworks, one made by a human and one made by an AI (or at least, the kinds of AI we have today, using the methods that today's AI systems use -- this isn't a simple chauvinism in favor of carbon over silicon as an underlying substrate), the AI image is simply worse, because (very briefly and roughly) human effort has intrinsic value, connecting with other humans has intrinsic value, the total historical and social context of an artwork has intrinsic value, etc. So it's perfectly fine for people to update their assessment of a given artwork when they learn more about its provenance.
This is a leap in logic that I see so often, and I don't quite get it. It almost seems so natural to them that they don't even notice it.
Because, given that human effort has intrinsic value, all context has intrinsic value, connecting with other humans has intrinsic value, etc. - which I subjectively agree with - it doesn't actually imply anything about the value of the artwork in question. Yes, effort has some ineffable intrinsic value - but, by no means, does that imbue the outcome of that effort with any value. Likewise with context.
In terms of human connection (and human effort as well, actually), pretty clearly any claims of some grid of pixels "connecting" the viewer to the placer can be applied just as easily to AI (of the modern sort, not scifi agentic AI) generated grids as to manually generated grids.
07mk
1d ago
I watched Moonraker for the first time a few years ago and, as someone who only watched the Brosnan & most of the Craig Bond films (along with Dr. No and Goldfinger), I was pleasantly surprised by just how camp it was. That they actually devoted time and effort a sappy romance montage of the primary henchman with metal teeth having a meet cute with some girl and falling in love with her, and then actually had that pay off, IIRC, by the henchman turning good thanks to the power of love at the climax, was pretty fantastic.
07mk
3d ago
This discussion makes me often think about Forrest Gump, where the titular character's mother is presented as being heroic for prostituting herself to the school superintendent in exchange for allowing Forrest, despite being officially tested as having something like 75 IQ, to attend classes with everyone else, because "he deserves the same education as any other kid" or something like that. The film also, of course, featured the same kid, who needed braces to walk, just one day suddenly becoming capable of running, not only like any other kid, but to a level enough to make him the star running back to what seemed like a high level college football team, despite having zero other football skills.
I'm always highly skeptical of the whole "we must manipulate fiction because fiction inevitably, implicitly, unconsciously manipulates people's beliefs about reality" crowd, but I think there may be a grain of truth in their claims.
07mk
3d ago
I doubt it. School quality being better than other states' doesn't imply producing better educated people than other states, it implies producing a better delta in educated status compared to other states, controlled for the children's potential ceiling, and my guess is that both the floor and the ceiling for children in Mississippi are lower than for most other states. It is also but one of many, MANY dimensions by which parents measure their likelihood of moving to the state, and my guess is that Mississippi has a lot of negatives in other very important dimensions. Furthermore, even if those weren't true, this is the kind of thing that would take at least a decade to see confirmation on any meaningful differences in output, which means even more time before people start moving in meaningful numbers, and that gives plenty of time for people in other states to find and come up with excuses for why the differences in output, as measured by the education level of public HS graduates, isn't due to Mississippi's specific methods of educating.
07mk
3d ago
The phenomenon that was written about it Coming Apart and its consequences...
As someone who was educated in semi-elite schools for most of my childhood/college, I recall the real kick in the teeth I felt in my 20s when I learned, through experience, that people who actually obeyed rules and put in honest effort into improving oneself was a rarity, rather than semi-common (still likely a minority or barely a majority in the schools I attended). People who grew up in even more elite institutions and then stayed only in elite institutions professionally, surrounded primarily by other people with similar experiences, just don't seem to have the capacity to understand just how dysfunctional vast swathes of society are, and how much of keeping society running is making sure their dysfunction doesn't cause too much damage. It seems like just another case of the apex fallacy, which seems endemic in the culture wars, including gender relations, race relations, and immigration.
Now, one possible point of hope there is that it's easier than ever before to see direct evidence of the actual lives of the actual people with whom one doesn't share an environment. I've seen people reference this with respect to the popularization of bodycam footage since they became near-ubiquitous among police forces post-Floyd. However, people - including myself - had foolish, naive, stupid, idiotic ideas about the proliferation of social media bringing people of different ideas and principles together, when, AFAICT, it has done the exact opposite. And generative AI adds a new wrinkle as well. After all, you can bring a horse to the water, but you can't make it drink. So I'm pretty pessimistic.
07mk
3d ago
I don't think those examples actually touch on the point of difficulty, which is convincing every state to copy Mississippi in terms of whatever it is they did that caused improvements.
07mk
4d ago
latency is not ~ever picoseconds to start with - a clock cycle is 1/4GHz = 1/4 nanosecond = 250 picoseconds, and nothing is faster than that.
So far. I suppose we'll hit physical limitations in terms of the length of the circuitry divided by C, and I don't know how the math would work out, but considering we're talking about future tech, it seems unwarranted to talk about the limitations of current tech. If we get this down to femtoseconds, even a 1000x slowdown is measured in picoseconds.
Growing up in the 90s, I recall some of my friends sharing tapes with recordings of songs from the radio and CDs with each other. But what I don't recognize is people calling them "mixtapes." That was something you might share specifically with your boy-/girlfriend as a way to convey your affection, and the exact specific order mattered. Tapes you shared with friends were just tapes that had songs on them, because merely having the song available for play - even with the delay it takes to rewind/fast forward the tape to the specific track - was something very valuable.
More options
Context Copy link