Perhaps I've been hanging around with Robby too long but today he brought up an interesting point.
At some point in the very near future, we are going to be able to create full on pie/slime/whatever videos with any female or male that we want. Complete with dialog or just about anything else we want. Right? It will look really good. It will look "real." We know that.

So, what will differentiate everyone's videos? He says "content." Meaning who can write and create the best videos that appeal to the most people in terms of the actual content and what he calls the "psychological content." That content that hits all the WAM "triggers." Should we all start practicing our writing skills? Our idea skills?

Thoughts?

You can call me,

Al

I'll give the same reply I did in a previous thread about AI WAM.

To me AI will never replace the real thing. Part of the appeal is knowing that there is someone experiencing the sensations and elements of the session.
Without a real person, it's just a soulless series of animated pictures.

I've seen some AI WAM that doesn't look terrible, but it is still always obvious and will never replace the real thing.

AI is incredibly. But I think knowing it's not real and actually didn't happen just makes it sad.
Yeah you can see your crush get messy in what ever you want but still doesn't make it feel special like the real thing.
Just knowing a girl went through the humiliation of a mess is actually the reason why I like it.
Ruining clothes, leaving smells etc is what WAM is about. With AI it's nothing just an image.
Just my point but definitely hope it isn't main stream on here.

I think the point is that you won't know its AI. It will look real and if the producer doesn't tell you its digital, you'll think a real person got hit by a pie, fell in the mud, was gunged.

I think yes being able to dream up exciting senarios will be a plus as will the ability to program the AI. There are guys over in the AI group doing programming way over my head. Their investing the time in learning this shit. The "producers" then will be the writers and programmers.

The big losers will be the WAM models. Why put up with divas when you can erase them if they misbehave?

It seems likely that AI will disrupt the entire porn industry and probably Hollywood as well. Once you can get the AI to spit out a video that is dialed in to your exact tastes, and you can get it instantly, why bother with a third party model or producer at all? It'll eventually look perfect and probably be completely interactive. Want shaving cream instead of whipped cream? Done. Want white pants? Done. Jeans? Done. Blonde? Brunette? Done. Done.

The same goes for Hollywood movies. Why pay $300 million when an AI will eventually create the next Marvel Movie in an instant? We could watch all 14 million possible outcomes of 'Avengers Endgame' if we wanted.

Once the technology reaches that point (and it will) I suspect that human acting will be relegated to stage shows and theater. Just a quaint niche for purists.

The technology is already there for text and chatbots. The pace of change on this stuff has been incredible. If I were an actor, news anchor, influencer, OnlyFans model, etc. I'd be saving my money.

This is a really good questions. I think it will make creators here more important than ever. But maybe I'm biased and haven't seen Ai wam smile as much

I'm truly dreading this point. AI WAM does absolutely nothing for me and I find it pretty unpleasantly uncanny. At least now, I can easily differentiate real from fake but once we get to the point that we can't, I don't really know what I'll do. Will I try to convince myself that AI WAM will work for me or will I just have to rely on older stuff that I know is authentic?

Like others have said, knowing that someone actually engaged in getting messy is a huge part of what makes WAM a turn on. Sure, someone could make an AI video of Taylor Swift getting absolutely drenched in thick gunge but knowing that she never did that and most likely never would kills it for me. Hopefully there will be enough of an audience and producers/models that feel the same way to keep putting out authentic content. AI or not, people are still going to enjoy getting messy and wanting to share it with others.

The one case for AI that I have found use for is improving actual photos from older content back when digital camera technology was more primitive. It isn't there yet for enhancing videos though. I can enjoy this content since while it is AI enhanced, the source material is real.

I could get into AI WAM. But another aspect to the 'authentic' experience of WAM is watching it live before your eyes, orchestrating it, making it happen with real people, taking part in it. As Marty DeBerghi in 'This is Spinal Tap' says "the sights, the sounds --- the smells". Watching it back knowing you were there.

There's no doubt that AI is going to reach a point it is capable of providing an experience whereby it achieves all of these things.

But, inevitable though all this appears, I notice we tend to assume that this experience will, in the natural course of things, be available to all of us. I do wonder whether it really will? Or might there exist a possible scenario where an increasingly privileged and wealthy few are able to attain this whilst millions of others will not and ultimately have to make do with the real thing?

At some point (maybe sooner than I think) it'll be good.

But now. It is wrong in every, every way. The wam just seems to fall off bare skin. The people's voices and eyes and moods are off.

The triggers for me are completely lacking. WAM is humiliating, erotic, stupid, and liberating. The current mud content looks like a weird spa ad.

AI content generally is fucking awful. I hate the ear r@pe AI voice ads on videos now.

I couldn't agree more @ NYCMuckpup

This again.

I want to agree with Robby, but producing anything of quality requires dedication, knowledge, and effort. The printing press has been around for hundreds of years but not everyone can write a good novel.

AI isn't for everyone, but nothing is for everyone. Not everyone reads stories or looks at cartoons, and I consider AI content to be a similar sort of thing. I produce photo-realistic animated cartoons of women getting slimed on a fantasy TV show. In terms of the subjects' reactions I consider it among the best AI content out there - according to one YouTube subscriber "it's the best gunge channel on the site, and it isn't even real" - and the concept is filling a niche that no 'real' producer is.

Any discussion of AI on the Messy forum will get the same "AI sucks" response from the same people every time. Perhaps a more fruitful discussion could be had in the AI WAM group. I feel that some (but not all) of the negative reaction to AI content is based on seeing a single poor example and judging it all the same. There is certainly a gulf between what AI is currently capable of, and what most people releasing AI WAM videos are actually releasing - there's a knowledge gap as many people start from scratch with nothing to learn from.

Maybe in future this gap will reduce, but I wouldn't bet on it - if anything, as the technology improves, I think the gap between 'pro' and 'amateur' will increase. And I don't think the future is people making AI videos featuring real people. For a start, it's illegal where I am and clearly immoral. The ability to make deepfakes has been around for a few years now and I've seen exactly one example of it for WAM.

People talk about realism without acknowledging the artifice of the 'real' videos they watch. To me, a lot of 'real' scripted WAM feels in many ways more inauthentic than what I produce with AI: the blank expressions, the deliberate and overly dramatic tilting back of the head ("look up!"), the "I don't know", the just sitting there not moving. Thinking about the nature of the relationship between the man filming and the girlfriend agreeing to be filmed for public consumption is distracting. The paddling pool in the untidy, dusty garage is distracting, The bad lighting and awkward acting is distracting.

There's good AI and bad AI, just as there's good 'real' and bad 'real'.

For me, I think there are 2 parts to the discussion that are separate and distinct. At least for me, there is a huge differentiation between a real messy experience and watching a video or looking at a photoshoot. There is no comparison between the two. Sure, I still enjoy produced content but the enjoy isn't the same as it used to be. On that basis, I tend to compare produced content and the potential of AI as the same. It doesn't matter if one is artificial.

Now, I've been involved in AI for a long time. I think there is an over estimation of what can currently be achieved with current technology and more importantly, the problems and hurdles to over come still.
Imaging is the easiest one to dissect. Commercial models have come a long way but if you are wanting to produce good quality wam content, you're not going to be able to achieve that due to content filtering and safety guardrails. Future models will most likely have the same controls in place which leaves you with Open models.
Open models have seen massive improvements over the last year but they're limited in number of areas. They've been trained on much smaller datasets lack many of the contexts needed for Wam.
This is where I've been focusing much of my work. It is possible to train these models and with time and effort you can get some very good results however there are still many limitations that I don't see being solved any time soon.
If you want a single image, then yes you can get a good result but if you want a consistent subject or a time series of images as the mess increases then you're not going to be able to achieve that yet. There will likely be newer models released over the next year that will help close this gap but it's going to be a while.

Video.. well. It has the same problems as imaging but this is compounded by the amount of compute resource that you need. Commercial models have come a long way but they're still in their infancy.
Local video... Developing well but the compute and memory needed is seriously restricting what is possible and even if that resource was available, the models just aren't good enough for anything other than a curiosity.
I've yet to see an AI WAM video that's made me think WOW. For me at least it's a curiosity that worth exploring only as learning tool. I'd go out on a limb and say it's going to 5 years minimum before we see anything that people would pay money for.

The short answer... Quit worrying, it's a long road ahead. Enjoy Wam for what it is. In the meantime, some AI images that does show not all AI content looks bad.

Exactly. The difference between "acting" and "simulating" might just be not having to watch a "real" woman with damp hair standing in front of a pie splattered wall and pretending she's not about to get hit with her 50th shaving cream pie of the day.

Yeah, the technology is not there yet but I don't see anybody saying it won't get there. This isn't warp drive. This is "just" software.

I find AI stuff interesting at best, but it doesn't arouse me at all.
I need to know it's a real person.

I'm not here to rain on anyone's parade, but I'd very respectfully disagree. These images look fine as previews, but even on my phone, once I open them, things are clearly *off.* They're not the horror shows from a few years ago, but there are still fundamental issues in the way that the images are generated that absolutely take me out of the experience.

I won't go into everything that makes/doesn't make a WAM clip work for me, but something that matters is verisimilitude. It needs to seem real. I'm willing to suspend my disbelief when watching a clip with mediocre CGI and green screen effects because that's part of the experience. I can't suspend my disbelief when someone's arm vanishes seamlessly into someone else's torso.

The thing that does really bug me about AI is that there are a lot of charlatans and snake oil salesmen out there promising impossible results, likely to hype the value of their companies. We were promised that scaling would solve everything, when some fundamental hallucination issues with the content remain.

Like someone trying to get in shape who keeps skipping leg day. The top gets better and better, but there's still something that's not quite where it needs to be. That isn't getting better. "Imagine where it will be in 5 years!" feels more and more like an empty promise when it comes to addressing fundamental issues with the outputs that have existed since the models came out.

Does it have a use? Sure. I'm not anti-AI by any means. But I'm growing more and more suspicious that there will still be a foundational "off-ness" to AI WAM products, no matter how much more polished the results look.

Lastly, the idea that we will be able to create Avengers-level content with our home computers "very soon" is ludicrous. These engines still require processing power, energy, raw materials, training data, and so much more. But hey, that's part of the sales pitch: "Now anyone can create amazing content! Ignore that we can only create 30 seconds of stuff on loop. And ignore the flaws, I betcha those bugs will be fixed *soon.*"

The scary thing about all this (and not just for WAM) is that soon you will not be able to believe any photo or video that you see. For proof of this witness a few weeks ago, you had well known TV presenters or politicians expounding the virtues of an obvious scam, on Facebook. Were it not for the subject matter and the personalites involved (depicted!) these videos could easily have been taken for real.

A crude example might be the video of Trump kissing Musk's feet. However the examples mentioned above were MUCH better than that.

It does not bode well for the future.

I think the sentiment of my post completely flew over your head. I'm quite pragmatic about AI and don't get caught up with doom mongering at least in relation to WAM. WAM is complex by it's nature. Fluid dynamics, textures etc are difficult. I don't see AI as a threat to WAM producers in the short to medium term. One off imaging perhaps but there is no temporal progression. Consistent characters, outfits and progressive poses are impossible right now.
Will it happen eventually? I'd be a fool to say no but it's got a long way to go.

Thanks for clarifying! I actually did track all of that, I just wasn't clear in my response. My first paragraph was mostly meant to respond to your point and saying that I would consider the images all that good (not horrifying, but still not necessarily good).

From there, though, I was speaking more in generalities, not pushing the point of what you'd said further. But that wasn't necessarily clear, apologies!

Nothing to add to the conversation, but I always love learning a new word

-JD

We are all assuming that access to AI will be as cheap and easy as now. If the business practices of streaming services are a model, then there might be a future where all AI generation gets so real it saturates everything, at which point it is all put behind paywalls and you pay more and more for less and less. This won't affect big businesses so much but it will hit casual users.