Video content
Here is a video in 3 parts which I created with a variety of AI tools. The source material were sets of photographs which I took of the doll "Audrey".
She is a 1/6 doll, and here she has a JIAOU body 09F, one of the white ones (I am not sure exactly which one, I never can distinguish them. But 09F means: full bust, fleshy thighs, a nice body for a woman around 40 years of age.
The video does not tell the back story, but here it is: Audrey - wearing her glasses - is a high-school teacher. Someone invited her to that club, where people can display themselves - and make a fool of themselves - by performing some pole dancing.
Now of course Audrey never has done this. But for some reason someone made her do this - maybe she lost a bet, maybe she wants to impress someone special... who knows.
But here she is, the high-school teacher, with glasses, in her mature conservative outfit with the white long but tight skirt. Well, she has a transparent blouse - this shows that there is more to her than the high-school teacher... :)
Well, the video begins with her being on the stage, at that pole, the music starts, and now she is supposed to dance.
And that she does. And she does it quite well!
And she actually likes it!
Well, I am not going to tell more of what happens in the video - you can watch the three-part video yourself.
Technology used
I used a lot of the OpenArt imge-to-video tools. This allowed to create start-and-end image sequences, so the overall sequence then looks very continuous.
Unfortunately, OpenArt does block NSFW prompts and images, so when it came to the actual NSFW content, other than the usual "normal" dancing, then I had to resort to other means - and I could no longer have those start-end-image sequences. Therefore, those NSFW parts are unfortunately disruptive, are composed of short 5-10 sec video snippets.
The AI models were not perfect. Often I did use WAN 2.2, and for the start-to-end sequences I used Seedance 1.0 lite.
In the videos I did leave the AI errors and artifacts. My goal was NOT to create the perfect video - that is currently only possible with a lot of time effort. But I wanted here to document the current state of the art in AI video creation technology. This way we can compare later how good the progress will be. And I anticipate that there will be huge improvements coming! All those errors will eventually be mitigated and resolved. So I am looking forward to make again such a video in a year or so.
Business case ?
Creating AI video snippets costs. The OpenArt costs are quite reasonable, 1 minute of video costs between $6 and $12 (low quality, no thrills, no pro mode), some even less at around $3 per minute. Overall the video is about 12 minutes, so you can estimate the costs. I needed to purchase several times those "credit packs", these alone cost $60. Adding some other subscription costs, I estimate that these three videos cost me now about $100 altogether.
Not quite cheap. And I have not even factored in the time I spent: these were basically 3 full days.
But this is still way cheaper than a blockbuster movie... and I see great potential for such videos!
I need now to find a way of selling such videos. There is a market for NSFW stuff, as the adult entertainment industry has demonstrated throughout the years.
Here and today I am making these videos available for free - as a documentation of what can be done, and also as a little advertisement for the OneSixthuals work. In the future I plan to sell such videos, with the occasional freebee of course.
I do hope you do enjoy these three videos!
And - be aware they are for mature audiences only - in particular the Part 3. You have been warned!
The Videos: "Audrey is Dancing"
Part 1:
Part 2:
Part 3:
