For Those Experimenting With AI

[Ticklefeather00]

Which AI are you using? Which is giving the best results?

I use mostly ideogram.ai for base images and then alter them with openart.ai and also gimp.

Just wondering what others are doing?

 

[U.N.Owen]

If I had the money to subscribe to multiple sites, I'd probably try more options, but having done free trials with a bunch of them, I'm sticking with Ideogram for now. It comes closer to capturing what I have in mind than any other that I've tried... and it's gotten noticeably better after recent upgrades at putting tickling in an image when you tell it to.

 

[Kalamos]

Out of curiosity I checked out Ideogram.
What stuck out for me was, uploading your own images is a premium feature.
So is, keeping you generations private - they are automatically set to public, if you don't subscribe.
All very puzzling.

 

[U.N.Owen]

Out of curiosity I checked out Ideogram.
What stuck out for me was, uploading your own images is a premium feature.
So is, keeping you generations private - they are automatically set to public, if you don't subscribe.
All very puzzling.

View attachment 1074187

AFAIK, none of the AI generators have full functionality without a subscription. But yes, those features are premium.

 

[Kalamos]

AFAIK, none of the AI generators have full functionality without a subscription. But yes, those features are premium.

Define "full functionality".

 

[U.N.Owen]

Define "full functionality".

Most sites don't let you make more than a handful of images without paying. So that's a pretty fundamental functionality issue.

It would undeniably be nice if Ideogram's membership tiers only affected the number of images you can make at a given time, and allowed things like private generation at the cheaper membership level, though.

 

[duderino84]

I get all my action with perchance

 

[Kalamos]

Most sites don't let you make more than a handful of images without paying. So that's a pretty fundamental functionality issue.

It would undeniably be nice if Ideogram's membership tiers only affected the number of images you can make at a given time, and allowed things like private generation at the cheaper membership level, though.

Ok.

 

[Maneo]

Images aside, there are a number of AI-based apps that let you interact with AI "companions" that are either predefined or ones that you can create. It is relatively easy to create a companion who is deathly ticklish and put into ticklish situations. The AI engine generates tickle scenarios and stories that can curl your toes. I use Kindroid. And AI is only in its infancy as a consumer product. The price will drop and the quality will improve at an astounding rate. Tickle videos of your choosing that you create (with you as the lee or ler star) are just around the corner. I hope TMF stays abreast of these developments.

 

[knight]

I've experimented with ideogram. But I prefer more sexual themes with my tickling, so the program I use most is sexy.ai. The SDXL workflow paired with the PrefectiousXL model usually give me the best results, but it's a very animated look. There's also a workflow that uses Pony checkpoints and Loras. Mixing and matching them can yield some interesting results.

 

[Kalamos]

I tried some generators where you can enter seeds and negative prompts.
But I guess I am old fashioned, and prefer those where I can manually enter pictures of my own production, and work on keyframes.
It feels "more real" than generating everything from scratch.
Prompt only, however sophisticated, feels too alien to me.

 

[Ticklefeather00]

I tried some generators where you can enter seeds and negative prompts.
But I guess I am old fashioned, and prefer those where I can manually enter pictures of my own production, and work on keyframes.
It feels "more real" than generating everything from scratch.
Prompt only, however sophisticated, feels too alien to me.

Openart.ai is very good for editing your. own images.

 

[felixhenry]

Which AI are you using? Which is giving the best results?

I use mostly ideogram.ai for base images and then alter them with openart.ai and also gimp.

Just wondering what others are doing?

I've been using Ideogram.ai a lot too—great for initial compositions! OpenArt's enhancement tools are solid, especially for refining facial features or textures. I also dabble with Leonardo.ai when I need more stylized outputs, and sometimes run edits through Photoshop if GIMP doesn't quite get it there. Have you tried combining outputs from different models yet? Curious how others layer their tools too.

 

[Kalamos]

Openart.ai is very good for editing your. own images.

Not enough free credit to do anything meaningful with my own images, though.

 

[Ticklefeather00]

Not enough free credit to do anything meaningful with my own images, though.

View attachment 1074565

No it is pricey but a very good tool for changing facial expressions, hair, clothing, undressing, changing breast size etc...

 

[Kalamos]

No it is pricey but a very good tool for changing facial expressions, hair, clothing, undressing, changing breast size etc...

I can usually do enough drawing/photo editing to change those myself, and then let other generators fill the gaps, as needed.
I seldom generate from absolute scratch nowadays; I used to employ Bing, but they started censoring more and more the prompts.
Plus, I would rather use something I've drawn myself or captured from my own videos as the starting point.
It would resonate more with me, than the generated contents I usually see, which often look very generic to me.

 

[U.N.Owen]

I'm not at all sure that Ideogram's upload feature would do what you're looking for. From my admittedly limited experimentation with that feature, it's more focused on generating new images based on a derived prompt from your upload rather than really basing generations on the actual image. (If you remix and turn the remix percentage up high, it will follow the uploaded image more closely - but it won't modify it so much as copy it with variations.) And the "canvas" image editor for uploaded images doesn't really seem to do anything Photoshop or Gimp couldn't do.

I like Ideogram because I can't draw at all, and it lets me create fairly complex images that follow detailed prompts fairly well. (I usually do only photographic-style images, since I prefer not to make anything that looks drawn when I can't draw.) But for an actual artist like yourself who wants AI as a subsidiary tool, Ideogram might not do what you want.