This weekly roundup thread is intended for all culture war posts. 'Culture war' is vaguely defined, but it basically means controversial issues that fall along set tribal lines. Arguments over culture war issues generate a lot of heat and little light, and few deeply entrenched people ever change their minds. This thread is for voicing opinions and analyzing the state of the discussion while trying to optimize for light over heat.
Optimistically, we think that engaging with people you disagree with is worth your time, and so is being nice! Pessimistically, there are many dynamics that can lead discussions on Culture War topics to become unproductive. There's a human tendency to divide along tribal lines, praising your ingroup and vilifying your outgroup - and if you think you find it easy to criticize your ingroup, then it may be that your outgroup is not who you think it is. Extremists with opposing positions can feed off each other, highlighting each other's worst points to justify their own angry rhetoric, which becomes in turn a new example of bad behavior for the other side to highlight.
We would like to avoid these negative dynamics. Accordingly, we ask that you do not use this thread for waging the Culture War. Examples of waging the Culture War:
-
Shaming.
-
Attempting to 'build consensus' or enforce ideological conformity.
-
Making sweeping generalizations to vilify a group you dislike.
-
Recruiting for a cause.
-
Posting links that could be summarized as 'Boo outgroup!' Basically, if your content is 'Can you believe what Those People did this week?' then you should either refrain from posting, or do some very patient work to contextualize and/or steel-man the relevant viewpoint.
In general, you should argue to understand, not to win. This thread is not territory to be claimed by one group or another; indeed, the aim is to have many different viewpoints represented here. Thus, we also ask that you follow some guidelines:
-
Speak plainly. Avoid sarcasm and mockery. When disagreeing with someone, state your objections explicitly.
-
Be as precise and charitable as you can. Don't paraphrase unflatteringly.
-
Don't imply that someone said something they did not say, even if you think it follows from what they said.
-
Write like everyone is reading and you want them to be included in the discussion.
On an ad hoc basis, the mods will try to compile a list of the best posts/comments from the previous week, posted in Quality Contribution threads and archived at /r/TheThread. You may nominate a comment for this list by clicking on 'report' at the bottom of the post and typing 'Actually a quality contribution' as the report reason.
Jump in the discussion.
No email address required.
Notes -
One of my favorite bands just took a bunch of AI accusations, I guess, and he wrote a somewhat-pissed Substack post. That lead singer doesn't often step into culture war stuff, but this was close enough, I think:
and goes on to say that fighting AI art in this way is fruitless:
I regret that the culture war is poking random people in a new way in the last couple of years, and I can't help but cynically laugh at it. Not to mention how short-sighted it is. In that post, the lead singer details how much of a pain it is to do graphic design for music, and videos, and other art, and he hates it. Imagine if you could get a machine to do it? Also, it actually lifts up people who do not have money and allows them to make art like the people who have money do. Look at this VEO 3 shitpost. Genuinely funny, and the production value would be insane if it was real, for a joke that probably wouldn't be worth it. But now, someone with some Gemini credits can make it. This increases the amount of people making things.
I'm not sure I have any real thesis for this post, but I haven't been very good at directing discussion for my own posts, so, reply to this anecdote in any way you see fit. I thought it was interesting, and a little sad.
Yes but artists are a holy protected class and anything that takes their jobs away is evil. Nevermind that it has been known for centuries that art is an extremely bad way to make a living and that cameras already caused a crisis in the art world that every sophomore art student has a postmodern fit about.
My view is opposing AI art is anti-humanist. For every artist that can produce something anyone wants to look at, you have perhaps 1000x as many people who see something in their mind's eye but they don't have the skill to render it. That thing, maybe even that stunningly beautiful thing, never sees the light of day and dies with them.
Rest assured, most people have nothing beautiful to render or interesting to write in the first place, so it's not like we have some insane well of cognitive surplus waiting to be tapped into. Even with amazing AI tools most people will never put out anything interesting. But the true intellects and creatives only have time to specialize in so few things right now and I look forward to any leverage AI tools give them.
EDIT: lol, I posted that VEO3 video to my Facebook timeline saying something about how even kings could not commission shitposts like this and two different libtards unfriended me over it because of how wrong-side-of-history it is to support this technology that puts artists out of business. Of all of the gray tribe stuff I post that gets me a bunch of unhinged leftist reactions, praising AI stuff was The Line.
This seems like a fundamental misunderstanding of how creation works, though. Good ideas arise from craft skill, innate talent plus long hours of practice honing your perceptive faculties and understanding of the medium.
Feel-good movies love ego-boosting scenes about the regular ol' Joe Schmoe whose genius idea puts all those snooty artists to shame. But in reality, there are no people who've only ever bothered to cook instant ramen, who also have genius ideas for a creative dish, and there are no one-finger piano plinkers who also have great ideas for an amazing symphony. Tyros will have either painfully conventional ideas that they don't realize are copies, or completely random ideas that add nothing. At most, in some rare cases, they might have natural inclination plus the germs of some concept that needs to be worked out through long years of development; so having that natural process short-circuited through easy access to AI slop will result in fewer good ideas ever seeing the light of day.
I guess the one exception might be niche porn as mentioned downthread, where each man knows best the precise configuration of tentacles, chains and peanut butter that will get him off. But that's less creativity than it is targeted stimulation.
I don't think this is a fundamental law of the universe, though. It's a result of the fact that a good idea is only good if it can be implemented in reality, and as such, people familiar with and talented at the craft of implementing ideas to reality - i.e. in the case of images, are skilled illustrators with lots of experience in manually illustrating images - are the ones able to come up with good ideas.
But as long as it results in a good image, the idea behind it is a "good idea," regardless of who came up with the idea or how. Now, people can translate ideas into images without that deep understanding of the medium*, with that translation process bypassing all/most of the skills and techniques that were traditionally required. And because of that bypassing, what constitutes a "good idea" no longer has the same limitations and requirements of being based on one's understanding of those traditional skills and techniques.
* Some may argue that diffusion models are a medium unto itself with its own set of skills to develop and practice, akin to how photography and painting both generate 2D images but are considered different mediums. I'm ignoring this point for now.
But this account leaves out the equally critical perceptive and analytic skills that are normally built side-by-side with physical skills as an artist practices their craft. The bare act of clicking a shutter is the same for me and for a pro photographer, but the pro will take an immeasurably better picture because they have a trained eye to compose it. I suspect they'll also take a better picture because they understand from long experience what are the strengths and weaknesses of that type of image, versus a painting or architecture, and can better choose their subjects in consequence.
I think part of the problem is using the same word, "idea," to describe both what goes through my casual-consumer mind and what goes through the mind of a trained artist when we think of a new image. The two are strictly different in informational content, but also in structure, as anyone can see for themselves if they scoot out from their Dunning-Kruger zone to consider an area of craft or creation where they are experts. Coding or software engineering are probably the most familiar arts for the Motte; when we're talking really elegant and well-built programs, is your uncle's "y'know I always thought we should have like an app for identifying hot dogs" the same as a technical concept that occurs to a high-level professional with years of practice? Is there anything shared between the two "ideas", beyond the inchoate consumer instinct "I want a thing to make me feel _____"?
I think a lot of speculation about the value of AI art relies on the stickiness of cultural premises from the pre-AI age, so when Joe says to ChatGPT "paint me, uh, a pretty elephant with an orange hat in the style of Monet" and gets some random pixels farted out using patterns from 10,000 human-painted images, we instinctively respond to the patterns with the delight we've learned to afford skilled human work. It may seem that we get that delight from Joe's "idea," but what we are actually enjoying is those other artists' artfully-constructed patterns. I don't think we can fairly expect that 40 years hence; I suspect people will just paw indifferently past most images the way we walk past tree leaves today, with the exception of any pics that happen to raise a boner.
Artistic skill-building requires a medium where you can exercise agency, though, because the agency or artfulness is fundamentally the part that we admire about it. For example, nobody looks at a Jackson Pollock painting and feels delight over how this black droplet aligns with this other black droplet, even though subtle visual details at that level are matter for praise in other painters. But things we know to be random or unintentional are generally not interesting, so instead fans enjoy Pollock's expressive choice of colors or line or concept, areas where he clearly did exercise artful choice.
With AI image generation, there are so many levels of randomness and frustrated choice that it's hard to imagine how a user could work for years to achieve progressively greater mastery. Don't most commercial models actively work to disrupt direct user control, e.g. by adding a system prompt you can't see and running even the words of your prompt through intermediate hidden LLM revisions before they even get to the image generator?
Commercial models are usually pretty limited in your control, but local models can be surprisingly deep in terms of technical skill.
There aren't many people working in the space yet, but there's a lot you can do. Inpainting allows controlled redrawing of selected areas, LoRAs (and, previously, Dreambooth) can be used to encode characters or things or styles or perspectives, Image Segmentation can control layout, ControlNet can be used to manipulate pose or composition, so on. Currently, first-frame-last-frame-packing video generation are pretty focused on something very akin to putting together a 'storyboard', and the most plausibly consistent that storyboard is drastically changes how consistent the output image can be. Local AIgen workflows can look very different from talking to a midjourney bot.
Some of these technical skills even have a little overlap: knowing things like the names of different paint or painting techniques, or how camera lenses work, or what poses people can actually do, or why composition matters, feed back into even prompting and heavily feeds into these more technical uses.
The big difference is that (with the arguable exception of storyboarding) these are technical skills; they'll show you how well you achieve what you're trying to do, without necessarily changing whether what you want to do looks good. Conventional artists always had a little bit of that -- drawing a circle or line to improve hand coordination doesn't inherently teach where to use those primitives -- but AIgen does not really have a good way to develop the skill of taste beyond personal preference.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
I oppose AI art because AI art (usually) gives money to AI companies (who are trying to end the world) and will at some (unknown) point become a memetic hazard to anyone who sees it. I think this is plenty humanist.
I agree with you about the "oh noes the artists" people, though.
More options
Context Copy link
In my experience so far, for every one AI-generated artpiece that was a genuine improvement over the alternative of "nothing" or "imagining it by reading a text meme", there are 10 thousand pieces of absolute slop that should have never been published with less effort than it took me to scroll past. I'm willing to take the tradeoff: a few true intellects publish a few less gems, in exchange for no more slop. We were not in danger of not having Enough Shit To See On The Internet as it was.
If I was AI regulation czar I'd consider the middle ground: you can generate all you want for personal use but you can't clog other people's eyeballs with it.
I see similar things on my social media, and I feel the exact opposite. The things that people call "AI slop" are, almost universally, things that would have been considered incredible works in the pre-generative AI era. Even today, they often have issues with things like hands, perspective, and lighting, and though they're often very easy to fix, just as often they aren't fixed before they're posted online. But even considering those issues, if someone came across such works in 2021, most people would find them quite aesthetically pleasing, if not beautiful.
So now we're inundated with this aesthetically pleasing slop that was generated and posted thoughtlessly by some lazy prompter, to the point that we've actually grown tired and bored of it. I see this as an absolute win, and I think my experience on the internet has become more pleasant and more beautiful because of it. I see it as akin to how Big Macs have become considered kind of slop food and eating it every day - an option almost anyone in the Western world has - would mark you as low status in many crowds, but for most of human existence, if you had that easy and cheap access to food that was that palatable and that nutritious, you'd be considered to be living an elite life. I think, for such access to such high quality food to have become so banal as to be considered slop is a sign of a great, prosperous world that is better than the alternative. So too for images (and video and music soon, hopefully).
I agree re: food. Not so with art. The entire purpose of it as I see it is for me to not be tired and bored, not to let me consume 1000 pictures of adequately technical, adequately colorful, adequately proportional and completely fungible content.
More options
Context Copy link
More options
Context Copy link
I'm into hentai games. Over the last couple of years, tons of titles have come out that use AI art, some of them quite good: Netorase Phone, NTR Phone, Fetish Phone, Blurring the Walls, Moonripple Lake, College of Mysteria, etc. I'm pretty sure the alternative to AI is not "the dev suddenly gits gud at drawing" or "the dev magically gets a huge art budget to commission illustrations", it's "the dev is reduced to reusing real porn clips" (for the phone games) or "the dev never makes the game in the first place" (for the visual novels).
Unless you reeeeeealy get off to this particular niche, there are tons of great games that already exist. So much so that you'll likely never get through your backlog in your lifetime.
There's a lot less 'adult' games out there than you might expect, even for a pretty shallow level of curation and taste.
F95 doesn't have literally every porn game (or even every recent porn game; Fek's Spellbound (cw:m/m) has a surprisingly interesting central mechanic and afaict no f95 thread), and its filtering tools aren't the best, but the widest tag search gives a little over 2k items for 'cheating'. I don't have a good grasp on netorare/netorase/netosi, but the whole category of cheating is probably not really what erwgv3g34's after, and a lot of games will get marked for a kink for having one or two scenes even mentioning it rather than any real serious focus. I'd be surprised if there's more than a couple hundred, and as much as the ones erwgv links are Bog Standard In Progress VNs, even moderately good writing or design would probably put them at the top 10% of the pack.
There's a lot of furry games (F95 gives about 1200), and we tend to be pretty omnivorous when it comes to material. But a lot of those are either tech demos, like most VR games or very-early-prototypes, or games with maybe a couple story paths or a handful of animations, like VNs or platformers. There are some games that are complex enough to be worth visiting a few different times, most of which I've either bought or donated for. But if you wanted a list (and, uh, were pretty open-minded), I'd still probably given dozens, not hundreds. Even of the one-shots, a pretty sizable portion are not great, even by the standards of LLMs or AIgen. You can run out pretty quick without being some supergooner.
That's most severe for adult media and for kink, because there's a lot of kinks that range from personal nonstarters even in tiny doses (eg, I personally don't like hyper period, no offense to those that do) to possibly illegal in many countries (ignoring the obvious examples, people have gotten pretty seriously legal concerns over producing a cartoony CBT game). Many of the better games for my interests are text-only, because the nature of what they are means producing more than a handful of paperdoll-style images is agonizing work even with a lot of artists willing to work for near-peanuts!
But it's not like it's specific to porn games.
A year or so ago Rayon brought a big conversation about Palworld. And it's never been that clear how much (if at all) AI was used in producing the game. But it's a genre with a mere handful of serious competitors, and even the most 'successful' ones are often much worse. Over a year later, Nintendo has brought a pretty deadly legal threat against Palworld; they don't have a game that can actually compete with Palworld's not-exactly-great implementation yet. Maybe Legends ZA will break the pattern in October, but the last Legends game was pretty much just a joke. ARK I remains notoriously unplayable for the vast majority of people, it has a 'remaster' that manages to be buggier, and ARK II is in ETA:soon mode and has been for three years now. Kenshi or Grounded, if you have a wide enough definition? There's a couple attempts in Roblox for dragon manager games, I guess? But that's an insult to gatcha.
Base-building survival game with critters seems like it'd be inundated with clones; it's not exactly the goats-on-fire of SFW gaming genres. And that's not really happened. AIgen isn't there, yet. Maybe those limitations are why nobody's been able to step into the space, and maybe it'll never get there. Or maybe AIgen will act as an attacter for the sorta people who make very shallow games. (after all, most AIgen on f95 is glorified VN)
More options
Context Copy link
There are tons of great games that already exist sure, there aren't necessarily tons of great games that align with a given person's preferences (and no, you don't have to be really into hentai to feel this way). I say this as someone who is not very interested in a large portion of the much-heralded games out there - there's an extreme deficit of games I would personally want to play. Everything that comes out of the AAA sphere may as well be slop as far as I'm concerned, since the approach that most large studios take when they construct games is basically diametrically opposed to mine. The increased output stemming from the democratisation of game development may well have resulted in an increase in low-effort content and a decrease in the average quality of games released, but the larger amount of content overall and the greater amount of indie games that are a product of one person's idiosyncratic vision has resulted in me finding far more games I enjoy. Arguably 100% of my favourite games only exist because of this process of democratisation, and I can't help but feel the same about the usage of AI tools to speed game production up and democratise it even further. I do not care at all about how the art was made; I only care about its ability to convey the intent of the developer behind it.
I once attempted to make a game on my own due to being unable to find anything I personally thought was interesting - making the art and animation was one of the most time-consuming parts for me since it is not my speciality, and I eventually had to resort to using preexisting photos and assets which I put through a heavy dithering effect and intense colour-grading in order to shorten development time. It would have been so much easier if I had done so in an era where AI tools were available to me. There's a shit ton of games made by inexperienced/time-poor developers with interesting ideas but where the stock assets are very visible; perhaps the existence of generative AI will reduce their incidence and encourage further creation.
I'm not too concerned about being drowned in low quality games; if one would prefer to avoid encountering slop entirely, there are many mechanisms that facilitate content curation and their importance and prominence will only increase as time goes on. It's not as if people are being forced to scroll through every shitty game that's been spewed out by an unknown developer in order to find something they like, that's a caricature that doesn't reflect the reality of how most people discover content; they typically find games through curation mechanisms like forums, review sites or recommendations by friends. Pointing to all the low quality content and wringing one's hands about the unimaginable horrors of All The Slop falls flat to me, since even in an overcrowded environment you can still effectively limit the scope of your search to a subset of media that's most likely to appeal to you.
Would you have liked all those indie hits if the artwork and text copy were noticeably AI generated?
In most cases I definitely would not have enjoyed them as much, no. Even I would say it would likely have taken away from the product - I do agree with you that AI art for the most part isn't inspiring to me, and there's a lot of noticeable artifacting in AI generations. Used as is, it's very immersion-breaking.
However, I'm not so sure it's likely to stay that way, and even in its current incarnation I can see very many use-cases for it. As an example there are many highly pixelated/low fidelity/dithered indie games which rely on the style precisely because of its simplicity, and it's not that difficult to selectively crop and edit AI image outputs in such a way where it's not recognisable as AI. You're still going to need to do a lot of work to make it look good and fit within the game's intended aesthetic, for sure, but it cuts down on time significantly when you're comparing against doing it by hand. Producing novel textures for 3d models are yet another possible situation where it could be quite helpful, I imagine. Its output usually isn't good enough to just use verbatim, but it can help speed up the process of game development and that's where I think its true utility lies at the moment.
More options
Context Copy link
Factorio would still be a banger. Creeper World would probably improve significantly.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
Aren't you describing usual power law stuff though (w.r.t. art, the top <1% is the best and the rest is generally ignorable)? Is the ratio that different from human generated content?
The power law actually is worse with ai. Terrible artists will mostly either get better or give up.
Ai slop grifters generate ai slop in less effort than it takes you to dislike, block, and scroll away. It will take a lot to get them to stop pressing the button.
More options
Context Copy link
I don't know about the ratio of technical quality. But as it stands right now, AI art is largely samey and even the best specimen (that I could identify, obviously) have the trademark sameyness and do not exceed the best human artwork.
Suppose you searched for a particular topic on a picture website, before AI boom it'd be a normal distribution from, say, 30% human ability to 99% (with the bottom tail cut off because the people who can only draw stickmen with a pencil usually don't publish them). After, we get a massive injection of 60%, and it's all in the same style.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link