We live in turbulent, highly interesting times. I just had the pleasure of thinking up a name for what the photo artist Chokebelt is actually doing. I think femdom AI artist sums it up quite well. Or: Femdom-AI-Artist. That means: Chokebelt uses artificial intelligence programs to create photos of dominant leather mistresses that don’t actually exist. The results speak for themselves. They are almost too good to be true. It’s fun to look at the pictures, but the whole thing also makes it clear that we can no longer rely on photos as sources of documentation of truths and facts.
Today, it is possible to create deceptively real-looking images on the computer. This became clear when photos of the Pope in a white hip puffer jacket appeared. Or photos of Donald Trump showing him being arrested by the police on the open street. These were well-made fakes. Created with AI programs. So this technology is not free of dangers. But that’s just a side note.
After all, Chokebelt is not about politics, but about the subject of leather mistresses. His approach is purely artistic. Chokebelt publishes his photos on X, formerly known as Twitter. He took the time to answer my questions and provided me with a whole bunch of great images exclusively. I show a small selection here. Have fun looking, marveling and reading.
Lady Sas: Hello Chokebelt, how come you post AI-generated photos of beautiful women with dominant charisma on X/Twitter?
Femdom AI artist Chokebelt: Ever since I can remember, I’ve been into women in leather and boots. I’ve been imagining my teacher in a black leather catsuit since I was 10 years old. I didn’t think Suzie Quatro was pretty, but her leather outfit blew me away.
I’ve been on Twitter for years and it’s become my favorite entertainment medium. I’m not that into the other social media outlets like Insta and Facebook. Lately, I’ve been noticing images on Twitter that were very close to my own fantasies and that appeared to have been made with Artificial Intelligence (AI) or Artificial Intelligence (AI). Two accounts in particular caught my eye, @fixationfcu and @blueisnowblack.
Then I thought, I’ll try that with artificial intelligence, too. And I knew Twitter.
Femdom AI artist Chokebelt in interview.
Lady Sas: What background in BDSM do you have yourself? What can you tell us about yourself?
Femdom AI Artist Chokebelt: My great loves are dominas, femdom, BDSM in general, leather, boots, bondage, breath control, here especially strangulation. Hence my somewhat obscure name. I have often thought about changing my nickname, because it sounds a bit off-putting, but now I’ve been using it for so long that I don’t want to change it anymore.
I don’t want to go into detail about my BDSM background here, but I’ll say this: I have lived out my preferences thoroughly. In Berlin you can do that.
Lady Sas: With which program do you generate the pictures? Take us along a little bit: What’s the process like from the idea to the finished image?
Chokebelt: I looked around, tried a bit. My PC is too old to make Stable Diffusion work, so I looked for online options. Dall-E was too expensive for me, Midjourney too complicated. I ended up at PlaygroundAI.com. There you can work with various stable diffusion models and additional filters. Even the free version is pretty good and for 10 € per month you can produce 2000 images per day, which I consider a bargain. I first used Playground AI with Stable Diffusion XL, then Stable Diffusion 1.5. In the last days I’m more back to SDXL after there are some very cool new filters for it. Besides that I use Promptchan.ai. The site doesn’t have nearly as many options as Playground, but there – unlike Playground – nudity and sex are also allowed. In addition, Promptchan has two very cool features: you can include faces of real people in the picture and you can use photos with poses of people as templates.
Before I get started, I think about what I want to see afterwards. I try – and I think this is the secret of my own surprising success – to tell a story with the pictures. Now, Twitter is conceivably unsuitable for telling a linear story. That’s why each picture should be able to stand on its own. There is a short caption for each image, and the viewer’s mental cinema does the rest.
Too good to be true. Images created by chokebelt.
About the first longer femdom story.
My first longer story is set in a society where women rule and is about a man who is sentenced to death for photographing women up their skirts. I describe from his point of view the reception at the execution center, his last night and finally the execution on the gallows. In fact, you don’t see or read anything explicit, the story ultimately only takes place in your head.
I have this story ready in my head and am now trying to produce the images for it.
First I steal a good prompt, that is an instruction to the program what it should do. You can find prompts everywhere, e.g. at Prompthero.com or directly in PlaygroundAI. If I like a prompt, I copy it and change the parameters until the image matches my expectations. Let’s say I find a picture of a young woman in a summer dress in a Greek village. Then I change the summer dress to leather dress, the Greek village to a Victorian saloon, and I’m getting closer. When all the settings are set, I click on „Generate“ and see what the program throws out at me. It’s often very surprising and sometimes more creative than I am. The fine tuning then takes time. I change the settings, the prompt, until I get the image I want.
My goal is to create as natural people, quasi photos, as possible. Anime and things like that don’t interest me as much, although that’s considerably easier. Nudity and explicit sexual acts ineress me too, of course, but not for publication. I don’t want to spread pornography, but I find it more exciting to stay discreet and to trigger the viewer’s head cinema. Everyone can then imagine what is under the leather.
About the time involved in AI image creation.
Lady Sas: How long does it take to complete a painting? From the idea to the finished result?
Femdom AI artist Chokebelt: It varies a lot. Sometimes an image fits from the beginning, sometimes I produce hundreds of images without getting the result I envision. It’s a lot of try and error. In addition, there is my fatal tendency to perfectionism. Actually, the images don’t have to be perfect in detail. They are usually only displayed on a small smartphone screen and only the overall impression counts. But I see mistakes in almost every picture. Today, for example, I produced 2000 images for 13 images I posted on Twitter. That’s a lot of garbage, of course, but also hundreds of wonderful images that just don’t fit the story. I could flood Twitter with my pictures.
After I set my preferences – the „prompt“ – the actual generating is actually very fast. I can select generating up to four different images at a time. After I click „Generate,“ the images appear after about one to three minutes.
Lady Sas: What inspires you about the images?
Femdom AI artist Chokebelt: I do mostly stuff that I like myself, of course, but I also try to bring in variety. The best AI stuff is coming out of Japan right now; the AI artists there are incredibly good. However, some of these artists have a very fixed style and end up doing the same thing over and over again. I want to avoid that. But a few themes run through my work, of course. Women and leather are practically always in the picture. I don’t have a fixed, recognizable style yet. Let’s see what the future brings.
About the problem areas in image creation with AI programs
Lady Sas: Where does your ambition to post something new every day come from?
Femdom AI artist Chokebelt: The sad truth? Fame Addiction. I delight in click counts, likes, comments. I stare at my phone for hours and am addicted to reactions. Before I started AI, I had between 100 to 300 followers, depending on whether Twitter had just kicked out the bots or not. Now I have almost 2000 followers. I’m afraid I’m wasting a lot of life time with my new hobby.
Lady Sas: Most of the time the AI camera is quite close to the women and shows them in the profile. Is there a reason for that?
Femdom AI artist Chokebelt: It’s mostly for technical reasons. AI still has problems with certain things. Faces that are far away – and a face is already far away when you do a full-body portrait – are often distorted, sometimes just a pile of pixels. When you have people sitting or lying down, the legs are often twisted, sometimes people get three or more legs. Controlling that is very tedious. And then you just leave the legs out. That’s often a pity, because I love boots. But you also have to remember that I’m a complete beginner. I started producing AI images just six weeks ago.
It doesn’t matter if the people are shown from the front or in profile. I haven’t noticed that most of my characters are in profile.
Lady Sas: I have heard that AI still has problems with hands. Six fingers are not uncommon. Where do you see the biggest difficulties at the moment?
Femdom AI artist Chokebelt: Hands are definitely a problem. That’s why a lot of AI artists – and me too – just leave the hands out. But I think this will get better in the next few years or even months. There are already filters now where hands are okay. In a pinch, you run the generation through as many times as you need to until the hands look good. Difficult are, as already mentioned, faces that are further away. I have not yet got a handle on multiple people in one image. Sometimes one person has two heads or several bodies are merged. It’s often a chamber of horrors. BDSM toys I also do not get right. Handcuffs, ropes or whips can not or only badly be represented. But leather looks really good now. Better than the original 😉
Femdom AI artist Chokebelt on the light and dark sides of artificial intelligence
Lady Sas: What do you think about AI in general? What is your position on it? Do you also see downsides and risks?
Femdom AI artist Chokebelt: I am completely thrilled by the possibilities of AI and am amazed every day at what AI makes possible for a rather artistically untalented person like me. The development is racing and we will see images indistinguishable from reality in the next few years. The possibility of transforming one’s own fantasies into visible images totally fascinates me.
Of course, there are also negative sides. Pictures will no longer have any probative value, but that’s not so dramatic. Even today, people are already lying with photos. The possibility of producing illegal content with AI is of course a problem. Graphic designers and photographers will have to adapt, and perhaps they will also become unemployed.
Overall, I’m optimistic about the future. AI is a wonderful toy that I’m having a lot of fun with. And I look forward to sharing my work with the Twitter community. Apparently, there are a lot of people who tick quite similarly to me. That’s wonderful, isn’t it?
Lady Sas: Thank you so much for your time and the many great images.
