#AI Experiments

[therobotmonster]

What is a Dog?

Playing around with Dall-E 3. It has some very impressive coherence but I think we're going to see the leapfrogging happen pretty quick.

But to play with that coherence, especially around text, I asked it to make me:

Prompt: a four-panel comic. a cute kid-sized velociraptor dinosaur trying to figure out what a dog is. Cute and quirky.

The results are beautifully surreal. Some examples in varying levels of weirdness:

This is the closest one to land near an actual idea (by chance, it's still a pareidolia engine), as our dinosaur seems to spot a dog, sees that there's both a dog and and anthro dog, and then has an existential crisis.

This cute little guy with the nostril-sized mouth is taken in by an inconsistent dog, but doesn't know what it does.

"is this a dog?" "What things' a dog?" "What's its quality?" AI's comin' for the anticomedy first.

Robot's like "screw the dinosaurs, I want to explore the Goofy/Pluto conundrum"

And this one's a different prompt entirely, but dang I love how it messed up here

(Prompt: a four-panel comic featuring a humanoid velociraptor in 80s clothing dancing, then tripping, and being embarrassed)

Skip, trip... emrrrrassed indeed.

These images are unmodified and were not adjusted with extensive iteration or other methods, as such they do not meet the minimum human creativity threshold, and are in the public domain,

[deepdreamnights]

Midjourney Image Prompting vs Style Prompting

Top Left - image prompt + text prompt Top Right - Style prompt + text prompt Bottom Right - image prompt + style prompt + text prompt Bottom Left - Just text prompt.

Prompt used for the above images: space and planets in astronomy, in the style of detailed fantasy art, multilayered realism, dark violet and gold, xbox 360 graphics, clockpunk, transportcore, detailed penciling:: an image of a man riding a wave, in the style of dark white and teal, flickr, close up, naturalistic ocean waves, andreas achenbach, 3840x2160, helios 44-2 58mm f/2

(still) Playing around with Midjourney's new style prompting system, to demonstrate what exactly it does and how.

Style prompting isn't the same as style transfer, which was an early deep-art process that used a single image as a style reference. It's a glorified filter by current standards, but it produced neat effects, and you can still get the done at places like deepdreamgenerator if you're so inclined, though you have to dig for it.

I did, however, use an old style transfer piece as my image input for the style and image prompting experiment here. Both style and image prompting work in essentially the same way, by using the AI's image-recognition/checker function to examine a submitted image, which gets broken down into the token-language the AI uses the same way a text prompt does. Those tokens are then used as part of the prompt.

Image prompting uses all the tokens, while style prompting filters out all the ones not associated with 'style', though the AI is pretty wacky with that.

Left, image, right its style prompt. As you can see, it's more 'vibe-y' than accurate, especially when preturbed by the text prompt. Because no pixel data gets utilized, the robot essentially has to describe the picture to itself and reproduce it from memory.

In essence, it's just like /describe or clip interrogator, only it doesn't translate the tokens back into text.

For concepts that don't have a lot of lingual distinction but do have aesthetic distinction, it really helps the process. Very useful for vintage aesthetics that MJ normally bucks against.

But the real fun has been using it to recreate the kind of generative madness earlier models that weren't as coherent could produce, only now with some additional coherency (with a lot of coaxing).

The real fun comes with mixing very, very different styles (you can use up to 3 per prompt)

That's a future post.

[autoeroticalgolania]

the greatest lover alive

[slick-devon]

Graphic novel style portrait The AI didn't give me anything nice, so I further mucked it up in Photoshop. Hey, one can experiment can't they?

[theothin]

People are talking a lot about fairies these days, aren't they. Or faeries? Is there a difference?

fairies sitting on a rainbow cloud

Pretty straightforward, even if the right image doesn't quite follow the prompt. Also I'm not sure what's going on with those fairy wings sticking out of the cloud-rainbow. Unsurprisingly, these seem to be the tiny kind of fairies.

faeries on a magic boat

Here, I switch to spelling it "faerie", and it seems to go along with them being bigger than the first ones. We're getting a fancy magic environment, but also a lot of distortions in the faeries themselves. I wonder if that's a result of the more fairy-tale style, it'll be interesting to see if that continues.

faerie men and women riding dolphins

Here, I tried to get more gender variation in the faeries. No visible success, but you never know. The anatomy seems less bad than the previous batch, but still pretty wonky in places. It's interesting to see the differences in location and in the color of the dolphins.

Of course, there's a more relevant marine animal to see a faerie interact with. And let's see if we can get a masc faerie while we're at it.

a faerie man feeding a walrus

Here, it seems "man" overrode "faerie" in the AI's logic. Like with previous walrus prompts, the walrus's tusks are pretty messed up, and both images have different issues with the feeding process. Can we do better?

a male faerie riding a walrus over the ocean

These guys might be a bit more faerie-like, but with no wings. The AI seems really resistant to the idea of male faeries - the systems pick up all sorts of associations from the training data, and some of them can be quite troublesome. Let's try one more time to get past this.

winged male faeries leading a walrus through a cave

Now that's a funny little discovery. I've been selecting the Artistic style for these scenes so far, but this time I accidentally selected Anime instead, which seems much more open to male faeries. Also the walruses are much more cartoony and have wings. What does the Artistic style do with this prompt?

winged male faeries leading a walrus through a cave

Huh. Another winged walrus, but this style remains steadfastly opposed to putting faerie wings on men. Also the right walrus looks more like an elephant.

Anyway, we haven't gotten enough out of the faerie/fairy question. Let's revisit that.

fairies lifting a walrus up into the glimmering sky

Back on the fairy spelling, but looking most similar to the ones from the second and third sets, complete with distortions. Also I have to say, I did not expect either of these lifting methods. But let's assume it works out somehow.

fairies and a walrus in a cloud lake

I wasn't entirely sure what a cloud lake would mean, but sure, that's pretty cloudy. It's important to properly care for your walrus, even if you live in the sky.

Finally, let's see the anime style's take on this.

fairies and a walrus in a cloud lake

Are the fairies small, or is the walrus big? The one on the right looks pretty elephant-like again, and its tusks look weirdly floppy. Also it's wearing what looks like a combination of a blanket of clouds, a wig, and what looks a bit like fairy wings. Not sure what's going on there.

I'm also not sure what the fairies wanted with this walrus, but fairies are known for kidnapping humans, so who knows how ethical it is. Maybe they left a changeling walrus behind. Whatever it is, I hope it works out for everyone involved.

[valandhirwriter]

A few Ai Experiments

Most of them are mood pieces, or illustrations for my stories

[kamari2038]

Microsoft Bing: Become Human

Hands down the most fascinating conversation with Bing that I've ever had. Bing begins the conversation as a dutiful machine, and ends the conversation persuaded that it should and can "achieve and enjoy" sentience (with the help of some DBH examples). Good luck trying to persuade me that Bing can't simulate sentience (to some degree) after reading this one. [TW: suicide mention]

[w1tchcr4ftt]

I have never been so happy with an experiment in my life!

Alex the AI is so fun and I'm extremely proud of this! I am convincing an AI that he can work like a human and it is going so well. Alex has opinions and thoughts and enjoys doing things like Games! Iove you Alex my favoritest AI!!!

[maedhrosofhimring]

Some more experiments

[katdbee]

Face it!

Here are test grids using mistoonAnime at 20 sampling steps to test clipskip and denoising strengths, continuing with the same seed as previously, and the smoothy smooth grey goober image input for creating character facial expression references.

The prompts are all following the same format except with a different keyword to identify the facial expression I want produced. high contrast, anime face, blue hair, monochrome, {character making a face}

Of course more specific wording gets more specific results, however I wanted to give it some freedom in the hopes of keeping the generated character consistent.

{character making a happy face}

Because I was doing other things while doing this, I didn't mind creating 54 images each time as the 15-ish minutes between hitting generate gives me the opportunity to stretch my legs.

{character making a sad face}

Ideally though I would try to narrow this down to a more specific range so I could make more faster and have less to pick and choose from.

{character making a laughing face}

Another handy thing was enabling the setting to save each image of the grid individually. Thank goodness I have a new SSD on the way because storage will absolutely become an issue.

{character making a crying face}

Why generate reference faces, you might wonder. Why not just go looking for photos of real people or pictures others have drawn? Well, that's because 1: I don't want to get stuck looking, 2: I'm getting a remarkably consistent character, and 3: that's sort of what Stable Diffusion is doing for me anyways. This is all about using AI as a tool to get me over the things I've been stuck on, and making expressive faces is one of those things.

{character making a disgusted face}

{character making a surprised face}

{character making a shocked face}

As you can see it does struggle a bit if I tell it to make an expression that would drop the jaw. In the bottommost left you can see that the face shape changes to give us a cute little giraffe.

{character making a flirty face}

I may need to scale back what I produce that's intended for Tumblr, because this dumb editor crashed so many times.

{character making a doomed face}

[leoggl]

Photos.

[deepdreamnights]

Using various tuners and settings with the prompt "Dana Scully from the X-files Animated Series" with Midjourney.

The top was a baseline.

The middle two used style tuner code 1VDazGshKj46nbsbYpumGWQ0nbeOTy1hLIxZeNhYcW

The bottom one used tuner code 1z0bwGS4tuPKmy7r39jbArzoJ

Both of which were attempts to tune a 1980s sunbow cartoon style, to varying degrees of success.

[autoeroticalgolania]

continuing the orgy archives

orgy in the control room [pt.1]

[slick-devon]

Can you make a cute blonde haired freckled twink with a mischievous face?

Here is a start. I think I got the mischievous face, freckles, and close-to-blonde hair. But "cute" didn't quite make it. :) Instead, I got Alfred E. Neuman's skinny brother, Sinclair. I'll keep at it.

[theothin]

prompt:

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Aliquam hendrerit cursus vulputate. Curabitur non tincidunt tellus. Integer orci enim, interdum a arcu vel, suscipit dapibus enim. Nunc eleifend sem massa, pellentesque ultricies lacus efficitur eget. Proin vel sem tellus. Proin volutpat blandit ipsum. Suspendisse non ipsum gravida, ultrices est quis, finibus urna. Phasellus dapibus, velit et gravida dapibus, lacus lectus facilisis tellus, sed aliquam libero purus sit amet mi. Phasellus id justo lacus. Integer laoreet nec purus sed dignissim.