A common complaint about AI generated scenes is the "uncanny valley" effect, where the scene is almost, but not quite, right, and it throws the viewer off slightly. TBH I expect this will get better - look at how far things have come in just the last two years - but comparing even the best AI pics and videos with actual human shot material, there's a huge gulf yet to cross. I'm interested in, and partly working in, the AI art field, but from my own POV I've found the most satisfying AI images are not near-realistic photos, but rather sumulated paintings, actual "art", that you might find hanging on the walls of a WAM-friendly stately home, or WAM-themed fine art gallery. Because paintings aren't meant to be photographs, things that look "odd" in AI photos aren't anything like as noticable in AI paintings.

My own favourite classic art period is the Victorian Pre-Raphaelite movement, artists like Edward Burne-Jones and his contemporaries, beautiful images full of detail, allegory, and visual storytelling. Of course in the 19th century painters didn't usually show women in overalls or tracksuits taking mudbaths in gothic bathing chambers, but by the powers of AI, I can now create those kinds of images, in something approximating the Pre-Raphaelite style. I've added a few of my favourite creations below.

Could this be an entirely new genre of WAM art, alongside WAM stories, as opposed to replacing human-produced scenes?

(for anyone who wants to experiment, the images below were procuced using PromptusAI's "Pre-Raphaelite-Art" microsite https://pre-raphaelite-art.promptus.ai/)

Also, I've not set the synthetic flag on this post as it's about non-photorealistic AI images - I'm interested to see what everyone thinks of this use of AI.

Very smart idea and comes a lot closer to being satisfactory to me. The lack of humanity therein is still somewhat gross, and not the fun sort of gross, especially when so much of your own work is all the better for the humanity involved.

It does open up the prospect for some fun jokes that I wouldn't have the talent to execute otherwise though, like a splosh remix of Andy Warhol's Campbell Soup cans, this time from the inside out.

WAM cartoons and comic strips have been a thing since the beginning.

But whatever the style, it's all a computer facsimile, and in a sense it's all a cartoon, regardless of how realistic it is. It's the representation of the situation that turns me on, and the representation has to be "good enough". Maybe some stylised 'anime style' stuff could work for me. A few years ago a couple of people were making some CGI animations and a couple of those were "good enough".

There are some 'non-realistic' images in the "a place to post ai content" megathread in the AI WAM group https://umd.net/groups/a-place-to-post-ai-content

Is John Everett Millais' Ophelia against TOS?

Don't worry, we are not going to be replacing modelled scenes with AI ones, Saturation Hall and Langstonedale will always be 100% real humans in our for-sale content. We might add some AI images of the house, estate, historic events aluded to in our backstory (if I ever have time to redevelop the sites), as decoration, just like how we have some random country house clip-art scattered about the place just now. But never for actual WAM scenes - if we ever did get into that, we'd create an all-new brand and sites for it, and make very clear which was which. But no plans to even think about that at the moment - still very much filming live, plus the huge stockpile we run with, also all live.


I remember reading somewhere years back about someone who'd created a very effective BDSM computer simulation that only showed triangles - but the story round it was sufficiently compelling it "worked" for the people it was aimed at. Any form of erotica is basically just about making brain suggestions - which is partly why there's such a huge range of WAM niches, and why almost anything will work for someone, somewhere, if it happens to triggers their specific receptors.


I'd forgotten about that, but yes, there are painting-style images there too.


The model was 19 when it was painted so I think that would be OK. I know she took ill after the lamps that were meant to warm the bath went out and Millais didn't notice, and she didn't complain so as not to spoil his work - but I gather she made a full recovery.

I was thinking more that it's a representation of someone dying.

"You like-a da juice?"

That would definitely be a no-no in an actual production, but honestly not sure in an artwork inspired by a fictional story. TBH we have two scenes Maude did by her own choice - she asked to do them - which were inspired by the Ophelia painting (as opposed to the actuial story behind it) and those have always been OK.


The idea here was to explore a different use of AI, not to try and produce actual images to compete with real ones (which current AI cant yet match, whatever the hype says) with an approach that's more akin to WAM stories, obviously fictional rather than pretending to be almost, but not quite, real. I know some people hate it but I don't think it's going away so I was wondering how it could be used to enhance, rather than compete with or detract from, the overall WAM ecosystem.


Not sure what that's a reference to but that's a nice image, thank you.

*** *** ***

Meanwhile, interesting to see the poll is skewing for realistic, rather than paintings. Fair enough, the objective was to see what peoople think overall.

Please tag this synthetic. I want to see actual images of actual messes, not slop from the plagarism machine.

Also why not ask in this dedicated AI wam forum where people might be interested?

The ToS specifically says that the Synthetic tag is only required when photorealistic imagery is posted, and that non-photorealistic work doesn't need it. https://umd.net/termsofservice/#deep_fakes hence why I've not used it here. The feedback is noted though.


That group is largely about photo-real synthetic images, with the attendant concerns about replacement of human producers and models, etc. In these paintings I was looking to explore an alternative use for AI in WAM that both resolves the uncanny valley effect, and expands the range of WAM media available, rather than threatening to replace human content creators. As a long established producer with multiple models who are also good friends, I have an interest in how all of this will develop - today I shot four wetlook scenes, and a spectacular messy one, with Lucia and Lily Kink, for example, something I very much want to be able to go on doing into the future. So as an idea that's at a bit of a tangent to most of what's posted in AI WAM, I thought Messy would be more appropriate to get a wider range of views and responses.

Forum posts are ephemeral - a week or two from now this one will have vanished into the depths never to be heard of again, as they all do eventually. Were I to post any of these paintings to my galleries, which are a lot more permanent, I would label synthetic and post to a dedicated (and labelled) gallery.

Interestingly the poll appears to be skewing in favour of photo-real AI rather than "third way" paintings, which wasn't what I expected, but was the whole point of polling it, to find out what people really think.

Meanwhile here are some teaser images for the five real wam scenes I shot today, something I very much care about being able to continue doing.

"You lika-a-da treacle juice?"

I agree with you about most things but here I disagree. The AI Wam group is exactly the right place for this post. The AI Wam group is about all aspects of AI in Wam, not just photo-realistic images, and you're not going to get all the "AI sucks" and "AI is replacing real producers" replies that add nothing interesting to the discussion.

i love that this site gives everyone a place to express options and ask provocative questions but i believe this poll is misleading. to be fair i would offer a 3rd choice. i do not care for AI in any format. i want my images to be real and i want to support those models and producers (especially those who consider themselves as amateurs) who care to deal in that media. not to pull her into my rant but God bless the mrs. b's that contribute to this site. well, she's retired but you get my point. p.s. keep getting better mrs. b..

This is a thread about AI. It literally has AI WAM as the first two words in the thread title. It's not a scientific poll and isn't intended to mislead anyone. If you're not interested in AI, you don't need to reply. Literally no-one is stopping you (or attempting to stop you) from buying scenes from models and producers who make real images.

I'm not interested in wetlook but I don't go onto the wetlook forum and say "I don't care for wetlook" on all the posts. DungeonMasterOne, this is what I meant when I said this post might be better off in the AI Wam group.

you put out a poll asking for opinions. i provided one. sorry if it's not what you are looking for but this is my opinion on this subject.

I didn't put out a poll. Your opinion is completely valid. It's just off-topic for this thread.

Edit: thank you for rephrasing your reply. As I say, there is nothing wrong with your opinion. But this is a poll asking about two different aesthetic choices with generative AI. We all already know that a significant proportion of this site's users don't like AI. And that's fine. But can we try to have a discussion and a poll about AI without insisting that "I don't like AI" needs to be one of the options?

Sorry I missed this, but this is exactly what I was getting at. The stuff I make is kind of realistic, but realism isn't really the end goal. It needs to look plausible but the end goal for me is a representation of the fantasy. The fantasy has "big mess" and involves a big element of ceremony and spectacle - getting that right is what's important to me, even if it veers into hyper-realism. With the real videos that I like it's the same thing - they set up a scenario and have a sense of ceremony, and they leave out the inconvenient aspects of reality that ruin the fantasy - I like scenes shot in plain studios and with no 'real life' stuff cluttering up the background.

Being a painter and illustrator by trade, this upsets me just as much as emulating photo realism. It's taking the soul out of thr equationor at best, torturing a machine to dance like a circus monkey until it produces a facsimile of scraped images of artists I chose to put lots of study into (as well as, possibly, some of my contemporaries) for this.

Beyond the material costs and ethical concerns, this is deeply threatening our ability to think through making.

Going outside of this designated AI forum and not tagging the fake images makes me want to avoid your actual content in the future. I'm sure it's great stuff but I don't want to support someone who's looking to further this harmful technology.
I realize that you're an established producer and you know the rules, I am not trying to police your actions or anything like that. I am telling you that I never want to see this slop, have used the tools the site provides to avoid it, and never go to the forum made to contain this crap, so seeing it breach containment is not appreciated.

Ok, that's a valid POV, and the first time I've seen anyone (here or in wider social media) mention thinking through making. Not something I'd ever considered before, but now you've said it, there is definitely something in creating for real that isn't there in creating with AI, it's the reason I don't think AI WAM is a serious threat to modelled WAM, the "spirit" is missing, at least so far. Point about illustration also noted.


The problem is you can't un-invent a technology - I'd love to live in a world without nuclear weapons but instead just have to accept that a slowly-but-steadily-growing number of countries posess arsenals that if ever used will end all life on earth in a heartbeat. So whether we like it or not, AI exists and is here to stay. The form may change, the companies almost certainly will - I'm old enough to remember when Compaq dominated PCs and Sun Microsystems were the last word in servers, both giants, now long-gone. But the tech will remain and continue to develop in both ability and availability. Which means those of us who want to remain in business need to be very much aware of it, it's strengths, and its limitations. Personally I've no desire to replace humans with AI, or I wouldn't have just blown half a month's revenue on a day of shoots with two models yesterday. But I need to know what I may be competing with, and also whether there are any non-threat uses of it, hence the idea in this thread. Though from the poll results, it seems most folk do want the photo-real versions. Which is still useful data, if not the answer I'd hoped for.


I think you're misunderstanding what Groups are for. MM has stated on many occasions that there's no requirement to post to groups, they are absolutely not intened to "contain" anything. If MM changes that then absolutely, but it's not how things are at the moment.

Regarding tagging: it used to be that a thread could be tagged synthetic or not, and all images, replies, etc, would inherit the tagging of the original post. The intent of tagging was to make clear when something was synthetically generated - so a thread that was just AI images should be tagged, but a thread that was discussing AI, but written by humans, should not. Plus, there was no requirement to tag anything that wasn't photorealisic, as if it's a drawiing or whatever it's clear it isn't real - same reason cartoon images don't need to be tagged.

However UMD is currently changing rapidly, with everything becoming much more granular, as this is needed for the coming roll-out of age verification and blocking, where it will be necessary to block access to different levels of non-PG content for non-age-verified users depending on where any given visitor is coming from - blame idiot politicians across the world for that. So a change that was made last week is that each image in a thread can now be tagged synthetic or not (along with R, X, etc), independently of the thread tag. I thought the "not needed for non-photo-real" still applied hence didn't tag the paintings, however MM has now done so - I'll need to check if that's a new policy for all AI images whether photo-real or not but will assume for the moment that it is.

From your POV this means you won't see the images again, but as discussions of AI by humans is allowed, untagged, anwhere it's on-topic, you can probably expect to see such threads, if relevant to messy content, in the messy forum from time to time. The only time an actual discussion thread would need to be tagged was if someone used something like ChatGPT to produce the actual text. But any associated images should be tagged if Ai generated, so those won't show.


Thing is I was hoping for input from people who may not visit the group as they think it's all photo-real imagery. And TBH all opinions, even negative ones, are useful data.


I did actually consider adding "AI is an abomination and all who use it should be taken out and shot" as a third option but thought that might detract from the data I was actually looking for. Am a bit disappointed that the poll is skewing much more to photo-real than artwork, so my hope that this might provide a "third way" approach isn't actually going to fly, but still good to get people's views.

To go along with this metaphor, you also don't need to be making more and better nuclear weapons. Most people aren't! If you don't want it to become a larger threat to your industry, then don't work towards improving it!

It's not a perfect metaphor - but anyone who has nuclear weapons is very much making better ones, while many others are trying to obtain them, the arms race will never stop.

Back on the subject of this thread, have had a chat with MM about it and policy and he's clarified:
The "Synthetic" tag is only to be used where the content is either AI-generated text, or photo-realistic images.
Discussions of AI by humans do not need to be tagged, and neither do non-realistic images, paintings, cartoons, etc.
And anything relevant to messy (or muddy) WAM is welcome in the Messy forum, there's no need or obligation to use groups, they are a convenience not a requirement.

He's taken the synth tag off the paintings above, had set it by mistake during a quick glance, TBH he's working flat out on the age verification stuff which involves major re-engineering of a lot of deep level stuff about how the site works at the moment, so easily done.

Personally, if posting photo-real I'd post them in the AI group as that's where people are most likely to appreciate them, but by definition that's a self-selecting audience.

Generative AI is not going away. It's precesly *because* I want to be able to go on shooting real scenes with real people that I need to keep up with the likely competition, both human and machine. On Monday evening I finally brought Miss Abigail and Lily Kink together in our dungeon for a couple of scenes to get to know each other. They hit it off beautifully and I very much want to bring them back for more, both mainline PG and Imperatrix Domina kinky. It's exactly because of things like that that I need to be fully up to date with the wider scene overall. Sticking one's head in the sand and hoping a new development goes away never ends well in business - see Kodak and the digital camera for a spectacular example.

PS: A couple of examples of what we got up to.

"The creme de la creme of the chess world
In a show with everything but Yul Brynner!|"

I know the last one is from the musical Chess - I love that song. Had to look up the first two, something to do with a US TV show, Saturday Night Live? Have heard of it but never seen it, so not sure what the reference actually means. Can you elaborate?

The thing about Kodak is that digital photography was higher quality and more convenient than film (at least for consumers) which is just not the case with AI videos. Plus there aren't any ethical concerns with film vs digital.
A better comparison is Quibi, who went all-in on something nobody wanted to spend money on and failed spectacularly.

Had never heard of Quibi, had to Google it. But that's missing the point. Kodak was a massive global company with enormous revenues from chemical print film, which invented digital photography but then stuck their heads in the sand and hoped it'd go away. And effectively threw away a multinational business.

As johnnypie has said a couple of times now, believing that AI won't get any better is like playing an arcade machine in 1982 and insisting that's the best video games will ever be. Those of us who want to go on creating need to be aware of and keep up with the developments, or we really will get replaced. This is the nature of things.

Personally my dabbling in Ai convinces me that as yet it isn't going to replace real-world WAM in the next 5 years. But that's an informed opinion having actively played with it. And while the poll ultimately went in favour of photo-real, I intend to continue with painting-style images from it, as I think that works better from my own POV.

Kodak failed because digital was better for the consumer, like I said. AI imagery is not, nobody wants it, like quibi. I didn't say anything about it stagnating. You have convinced yourself that as a creator, working with the plagarism machine is good somehow, which is baffling. You can keep an eye on the "competition" without espousing its virtues.

When you say "nobody wants it", that simply is not true.

I'm all for people hating on AI, and I agree that most of the results people get with it are shit. But I quite like what I do with AI, and I know that there are a few others who like it as well. Some people want it.

Ultimately, in our context, it's a toy for making low quality wank material. It's not stopping anyone else having any fun, it's not stopping anyone shooting video, or buying videos of real people. There are real problems and genuine concerns to have about AI in general, but when it comes to making this sort of simple 'erotic' content, I really don't see what the fuss is about.

Most or all of the people who like what I make would prefer the same thing shot for real with real people in front of the camera. DungeonMasterOne has been very clear and gone out of his way to explain that him mucking about with AI images isn't stopping him shooting real scenes and real sets.

It's really simple: we can have both.

As a producer, I have vowed to never use AI for my productions and will always hire models. I always want to support them. But that could also mean that I'm vowing to keep my "Fotomat business running" upon the advent of digital cameras. Over time, AI video creators making the same fetish videos at a fraction of my costs, could put me out of business.

There are some big assumptions here:

that AI video will inevitably get good enough to produce a ten minute long scene with good continuity. Will this even be possible? If so, what's your timescale for this? 6 months? A year? 5 years?

that AI video will be cheaper to produce than real video. I know that people can pay OpenAI a few bucks and make a few 8 second long clips, but to do AI even now as a hobby is expensive if you want to own your own gear. A GPU that can do wan2.2 decently costs about $8k. The GPUs to run the newer wan2.5 would cost about $60k. And hiring GPU time is cheaper but by no means free. When I briefly sold a few AI scenes, I sold them at a net loss. Why do you assume this will get cheaper?

that the AI video in the future will be better than what can be produced with real people. Why do you assume this? What could AI do that could be better than a video featuring a real person?

that the two can't co-exist. Why wouldn't the same people buy both AI and real scenes if both of them appealed? Why would it have to be one or the other?