AI-generated photo and video

[louislouis]

Agree with all above and in spite of what I said some days ago, have gotten into using Grok. l've found 'seeding' or 'priming the pump' works for me better than having it generate an image. I upload pics with wet spots, wet areas, and then have it generate a video, guided by building descriptions, again non sexual usually. Works quite well, if sometimes weird results that AI always gives. Some examples here.
https://w-si.link/MBGqflaT8K8n3HwCi
And not uploaded the vid but a work in progress.
Lots and lots more to come.....

Screenshot 2026-01-13 at 07.35.35.jpgScreenshot 2026-01-13 at 07.39.07.jpg

You've done pretty good with those videos, the first one where she does a little pee dance and the rivulets of pee running down her legs is best I think.

[Lawrence]

aqua 2021 I sent you a PM

[Lawrence]

The wait is too long for this girl

generated_videodh01a.jpg



generated_videodh01b.jpg

Link to the six-second video:

https://w-si.link/mv32kvRfpYrCZ5O0g

[Lawrence]

l've found 'seeding' or 'priming the pump' works for me better than having it generate an image. I upload pics with wet spots, wet areas, and then have it generate a video, guided by building descriptions, again non sexual usually. Works quite well, if sometimes weird results that AI always gives. Some examples here.
https://w-si.link/MBGqflaT8K8n3HwCi

I have found that I can use Grok to generate images of a girl with wet pants, and then use those images to generate video. Image generation doesn't have a lot of blocks and restrictions on this stuff, particularly when Grok is generating the image on its own and not modifying an image you have uploaded. Somehow, when you build a video, the guardrails may be less restrictive and less effective if the video is generated from an image that Grok created, particularly if Grok created the image from nothing, i.e., Grok "knows" that the image is not a real human.

Here's my latest work, using this particular method:

bus_stop.jpg


bus_stop_02.jpg

Link to the six-second video, with audio of the phone conversation:

https://w-si.link/XxZ1EFYO6U16uiAdA

[louislouis]

! have created some public wetting accidents of more mature women on the photo board, go over and have a look if you're interested.

[Lawrence]

! have created some public wetting accidents of more mature women on the photo board, go over and have a look if you're interested.

You're definitely into the MILF vibe, Dude

It's different for me. When I was 25, I had a vanilla relationship with a woman who was 41. It was awesome while it lasted.

Now I'm 57, and I still want to date a 41-year old LMAO

In real life, I would date a woman of almost any age, into her sixties if everything else is a good fit. I gotta be realistic.

For fantasy material, my upper bound is about 45 LOL

[Lawrence]

Now experimenting a bit with bedwetting. No good quality videos yet.

Genuine bedwetting videos and photos are, in my opinion, extremely rare. Adult bedwetting is very common, especially when alcohol is involved. But capturing it on video is rare, because most of it happens under a blanket. And even when it is exposed, then you need someone to take the picture or the video. And then post it. LOL



09_bw.jpg


10_bw.jpg


16_bw.jpg


18_bw.jpg


20_bw.jpg

[Lawrence]

And now for something completely different:

260120_acq_01.jpg


260120_acq_02.jpg


260120_acq_03.jpg

And a six-second video, with audio:

https://www.wesendit.com/dl/9EykPEJdfY7Ivyqru

Yes, the theme and the script get to be a bit repetitive. But there is only so much you can do with six seconds. Not sure yet if longer videos are possible. And when you try to make the action or the dialogue more complicated, that's when things start go awry, and the AI can't figure out what you really want. In making this video, I got multiple interations in which the girl was holding a cell phone in each hand--one phone at each ear. Probably because my prompt contained the phrase mobile phone more than once: I used the phrase once to put it in her hand, and once to have her talking into it. And AI doesn't always recognize that multiple instances of the same phrase may have the same referent.

Just adding a second person to the dialogue has to be done very carefully, or you get unexpected results, or complete garbage.

Feedback and suggestions are welcome. If you have concrete ideas, feel free to share and I'll see if I can make it happen.

I think this is the first video I have created that has standing movement that really works, with an authentic wet pattern, and view of the front and the rear.

[aqua 2021]

I've been working on my vids for days, and some great results, but tardy in posting; I will...this is very good And btw I have some bed wetting pics, real and AI'
d and Photoshoped, and happy to post those making clear which is which. more soon....

[louislouis]

You're definitely into the MILF vibe, Dude

It's different for me. When I was 25, I had a vanilla relationship with a woman who was 41. It was awesome while it lasted.

Now I'm 57, and I still want to date a 41-year old LMAO

In real life, I would date a woman of almost any age, into her sixties if everything else is a good fit. I gotta be realistic.

For fantasy material, my upper bound is about 45 LOL

LOL could maybe go as young as young as 41 The great thing about AI is you can create the fantasies out of your own head, before we all looked at the same photos and videos that had desperation and wetting but the girl/woman might have been too young too old, not wearing the right clothes or the acting may not have been convincing and so on. Having said that I still like watching videos of women just wetting them selves on purpose. I am liking yours and Aquas work though, keep up the good work.

[Lawrence]

Having said that I still like watching videos of women just wetting them selves on purpose. I am liking yours and Aquas work though, keep up the good work.

I am generally not into deliberate wetting. However....

I have talked about this a couple times in other threads. There is a gray area in which genuine, unplanned desperation turns into a controlled deliberate wetting, because the girl knows she will not get to a toilet in time and chooses to take control of the situation, and wet her pants on her own terms, instead of waiting until she has completely lost control in a way that would be more embarrassing or more difficult to manage.

So with that being said, you may like something that I created a couple days ago, by accident (pun intended), while working on something else that wasn't even wetting-related. I just saw an opportunity and decided to try something. And it's not a masterpiece LOL. The fluid behavior in the video is not quite realistic, so it requires a bit of imagination.

Libertine, body-positive women can get away with a lot while wearing a long skirt. But it takes some finesse, and things don't always work out perfectly.

Hippie chick looks like she's higher than a kite LMAO

img_ls01.jpg


img_ls02.jpg

And the video:

https://www.wesendit.com/dl/bPFc7BpFdrQ0J8bEW

[Lawrence]

For this one, I have not yet been able to generate a complete video that is watchable. I can't get Grok to generate dialogue that sounds lifelike and realistic for this particular context. I may continue working on it. I may not.

But the fluid dynamics may be the best result I've gotten so far. This may be the first video I have created of a standing accident in progress that almost looks real.


001_260124_arrest.jpg


002_260124_arrest.jpg

003_260124_arrest.jpg

004_260124_arrest.jpg

005_260124_arrest.jpg

[Lawrence]

... and here is the video, finally, with some dialogue that isn't complete garbage:

https://www.wesendit.com/dl/tX4OPD4QNafAthQY2

As always, feedback and suggestions are welcome.

For this video, I began with an image in Grok, and then once I had something I liked, I asked it to make a video out of the image. Then I had to go through dozens of interations of the video, tweaking the prompts for dialogue and other variables, including the behavior of the pee on her jeans.

I did not keep a record of the prompt for the image. It was something like this:

A 19-year old woman with long blonde hair is wearing loose light blue jeans and a gray tank top. She has been arrested in the parking lot of a liquor store by two police officers. A small wet spot is visible in the crotch of her jeans.

That kind of prompt did not encounter any blocks or guardrails for an image, and it generates dozens, even hundreds of images. And most of them are junk. You have to hunt for one that looks good. The video was more challenging, and I encountered a lot of moderation and blocks.

My prompt did not include any details about the police officers. No gender. No race. No age. No description of the uniform. No name for the police department or the city. Nothing. Nada. Zip. Grok generated those two cops entirely from its training data. And I made no effort to modify what they look like, in the photo or in the video.

Does the bald cop look like Hank Schrader, played by Dean Norris, in Breaking Bad?

That character was a DEA agent, and he seldom if ever wore a uniform that looked like the one in these images. He was not a uniformed police officer. He was a federal agent who either wore business attire or jeans and a polo shirt. Or, if he was on a tactical operation, he wore a combat-style uniform (like ICE agents in Minneapolis, but no mask).

Nevertheless, the resemblance is striking.

[aqua 2021]

Some stills. Vids will come...some really good ones, hang in... Screenshot 2025-10-29 at 01.26.55.jpgScreenshot 2025-10-29 at 01.26.47.jpgScreenshot 2025-09-03 at 20.57.11.jpgScreenshot 2025-10-29 at 01.27.05.jpg

[aqua 2021]

the police arrest vid is great, well done....of course given ICE at the moment (not the police, I get it) in the U.S. perhaps arrest vids are not the best...so many other possibilities.