Redjak's finally started using AI

[Boggy Man]

She certainly does look to be enjoying herself... at least until she realizes that her thick coating of mud is dragging her down...

Oh...

That's a spin on it that I didn't even think of Jason. No peril. No threat. Just pure pleasure.

But now that you put that thought in my mind...

Maybe I should try and create something along the lines you've hinted at?

A mudding session that goes wrong and shifts from pleasure to peril.

Hmmm...

A natural setup is where the mud may have a false bottom, which is a layer of stiffer mud that makes the sinker believe that it is shallow and safe , until the "bottom" starts to give way or liquify , and she then starts sinking again, without stopping!

Of course a "Redjak-engineered" pit could have a trap door, perhaps iris-style, under the mud, making it seem to be only breast deep, giving the victim a false sense of security, allowing her to have a "pleasurable" time , until Redjak secretly works the controls to open the iris "bottom" to expose bottomless muck underneath, and the real fun begins!

[redjak6t4]

Great ideas, Boggy Man. Thanks for the input.


For the natural set up to work a series of images would be needed. At the very minimum, two images. A 'before' and an 'after'. Starting with the damsel believing that she is safe and then discovering that what she thought was a safe mud pit... isn't.

The Redjak-engineered pit idea sounds good too. It would probably have to follow the same general sequence. Before and then after. If a kind of cross-section of the mud pit could be shown so that we can see the damsel's feet resting on the sub-surface iris, that might work.

If I were to explore either of these ideas BM I'd probably have to do so using photoshop. Grok is just fine when it comes to generating one-off images, but it does have it's limitations. It won't reproduce exactly the same image, time and time again. Instead it always varies the foreground, the background, the lighting, the pose, etc. So I can't rely on it to give me a before and after image of the same damsel in the same pose as seen from the same viewpoint.

Nevertheless, I shall file your ideas away in the darkest recesses of my mind and let my subconscious id think about them.

Who knows? Something new and fresh and dastardly may come out this.


Redjak.

[Boggy Man]

Here is yet another idea for a mudpit with a false bottom! How about the playable, irresistable-feeling mud sitting on top of a bottomless pit of tar that is so thick that it holds together and stays under the mud even though it is ultimately lighter. Because the mud on top of it is wet, the surface of the tar "bottom" isn't sticky, but feels like a regular bottom that has some softness to it. The mud itself feels soooooo good, that the female victims pleasuring themselves in it don't realize their feet are sinking into the bottom until it is too late, and the tar has closed over the tops of their feet! Any attempts to lift one foot out plunges the other in deeper, and the mud on top gives way when they try pushing down with their arms! Eventually, they vanish beneath the mud, and then finally, get sucked under the "bottom" beneath the mud, leaving behind no trace of them in the mudpit!

[redjak6t4]

Another excellent idea, Boggy Man.

Please keep them coming... but also please be advised that I can't possibly keep up with you.

If I do take one of your fiendish ideas and run with it, that'll be because I can see, in my mind's eye, how I might be able to make it work as an image or series of images. As I said in my latest reply to you, I have two strings to my bow. Photoshop and Grok. We know that AI systems like Grok can rapidly churn out huge numbers of images. But, because of the vagaries of AI, its sometimes takes hundreds of iterations to get anywhere close to the desired result.

I've already tried taking a suitable Grok image, altering it with Photoshop and then feeding it back into Grok. Unfortunately, Grok either downgrades the clarity and resolution of that image, rendering cartoon-like or less lifelike, or it goes off on one of it's unasked for tangents and starts doing what IT thinks I mean and not what I've instructed it to do. No amount of corrective prompts can then rescue the project. I then have to abandon it and start over.

Needless to say, this is very frustrating and also time consuming.

I suppose the accusation that some of my work is 'samey' derives from this kind of problem with Grok. By a process of trial and error I've discovered a good pathway to creating believable quicksand and muddy fakes, using a set of prompts that work reliably. The downside is that what I create therefore follows certain trends and themes and isn't open to much variation or imagination. Grok does a certain thing well, but doesn't do other things so well.

I therefore have to devote a lot of time and energy in experimentation, with many more failures than successes.

And this feeds back into what I was saying above about your creativity and new ideas outpacing my ability to bring them to fruition. Rest assured Boggy Man, that I'm really, really pleased with the feedback I'm getting in this thread. I just feel somewhat embarrassed and frustrated that I can't act quickly on your ideas and convert them into enjoyable images as quickly as I'd like or probably as quickly as you'd like.

I hope you understand.


With thanks,

Redjak.

[redjak6t4]

Here you are Boggy Man.

WAM 1.png

WAM 2.png

WAM 3.png

WAM 4.png

WAM 5.png

WAM 6.png

These illustrate the point that I made earlier. Once I find a reliable pathway (set of prompts) that Grok understands then I can just let it produce more and more images, only needing to select the best and then discarding the rest. The half dozen you see here were created in about two hours, with possibly fifty not making the cut. Grok created each one in seconds, but most of that two hours was spent experimenting with and refining the prompts so that it gave me what I wanted and not what it thought I wanted.

Mind you, its just a machine and doesn't 'think' at all.

Anyway, these follow on from my Mud Queen effort a few days ago. I consider them to me more WAMish than quicksand focused. Hopefully the folks here will like them.

Thanks,

Redjak.

[Boggy Man]

Another excellent idea, Boggy Man.

Please keep them coming... but also please be advised that I can't possibly keep up with you.

If I do take one of your fiendish ideas and run with it, that'll be because I can see, in my mind's eye, how I might be able to make it work as an image or series of images. As I said in my latest reply to you, I have two strings to my bow. Photoshop and Grok. We know that AI systems like Grok can rapidly churn out huge numbers of images. But, because of the vagaries of AI, its sometimes takes hundreds of iterations to get anywhere close to the desired result.

I've already tried taking a suitable Grok image, altering it with Photoshop and then feeding it back into Grok. Unfortunately, Grok either downgrades the clarity and resolution of that image, rendering cartoon-like or less lifelike, or it goes off on one of it's unasked for tangents and starts doing what IT thinks I mean and not what I've instructed it to do. No amount of corrective prompts can then rescue the project. I then have to abandon it and start over.

Needless to say, this is very frustrating and also time consuming.

I suppose the accusation that some of my work is 'samey' derives from this kind of problem with Grok. By a process of trial and error I've discovered a good pathway to creating believable quicksand and muddy fakes, using a set of prompts that work reliably. The downside is that what I create therefore follows certain trends and themes and isn't open to much variation or imagination. Grok does a certain thing well, but doesn't do other things so well.

I therefore have to devote a lot of time and energy in experimentation, with many more failures than successes.

And this feeds back into what I was saying above about your creativity and new ideas outpacing my ability to bring them to fruition. Rest assured Boggy Man, that I'm really, really pleased with the feedback I'm getting in this thread. I just feel somewhat embarrassed and frustrated that I can't act quickly on your ideas and convert them into enjoyable images as quickly as I'd like or probably as quickly as you'd like.

I hope you understand.


With thanks,

Redjak.

I do completely understand. I did not intend to pressure you in any way with the number of ideas I have been suggesting, and am sorry if that is what happened. When the scenario of a muddy session going from pleasure to peril was suggested, I just couldn't resist spilling out a bunch of ideas on how that type of situation could come to be, adding to your (Redjak's) arsenal of potential trap plans that could possibly be implemented in the future, if you decided to, and if it was feasible.

That being said, one final example of that scenario would be a submerged stiff layer that not only eventually gives way beneath the victim's feet to bottomless muck below, but automatically reseals itself after the victim has completely sunk below that layer so that it is a flat stiff false bottom again several feet below the surface!

[redjak6t4]

That's ok, Boggy Man. We're cool.

There's no need to apologise and my offer still stands... please keep the ideas coming. One of the delights in sharing a fetish like this is reading the thoughts of like-minded people who have clearly been thinking at length and in depth (pun intended) about it.

As you know, for years now I've stepped into the shoes of the deranged madman called Redjak the Quicksand Killer to create images of his nefarious misdeeds. But you've suggested something new here. Instead of him going out and acquiring victims to sink in quicksand, there's another way he can fulfil his twisted urges.

By setting up devious traps and tricks to catch the unwary female sinker. Suddenly turning her pleasurable frolicking in deep mud into a life or death struggle that is heavily weighted against her. This new and fresh angle is very welcome and even if I can't convert it into an image right now I can certainly see what I can come up with.

Who knows? Sometimes Grok surprises me by reading my prompts in a way that I didn't expect. Or reading them exactly right and delivering nice, juicy muddy images. So I really can't say with any certainty that I CAN or I CAN'T follow your ideas. We'll just have to see.

But as mentioned before, please keep thinking.

Thanks,

Redjak.