🤯 Any other AI art addicts here? 🤯

Lately I’ve been playing around with computer generated art and it’s really addicting.
With the latest release of Stable algorithm, it’s looking even better than before.

It takes a lot of tries and the results are usually not what I wanted but from time to time miracles happen and it comes out ever more awesome than expected.

I made this one last night and omc, it sure derailed all other plans I had for the evening.

This is Beira, the Queen of Winter, as created by NightCafe using stable and just a prompt. Definitely the best face AI has created for me so far.

I later touched it up in an editor and added realistic eyes to it. Gorgeous results! I’m working on creating a cover for it. I’ll post it once it’s done.

I also gave Midjourney a try last night since I’ve heard a lot about it. I ran out of the free trial but before I did, it created this awesome thing for me.

:scream: Merlin’s workshop came out even better than I imagined. I might actually change the description in the book to match closer to this. I’m in love.

Another one I’ve played with a lot is Dream by Wombo. This one doesn’t have limits :grin: but it doesn’t get to high resolution. Don’t try faces on it. Omc, it creates monster faces, but it’s really good at abstract art and settings.

So here’s Glass Garden City as imagined by Dream.

It’s a really cool toy. And did I say free?

I’ve also used Dall-e mini and Craiyon. They’re cool to play with but results are low quality.

Another cool one is Portrait style transfer with DualStyleGAN. This one creates faces. You supply an image and it creates a new face out of it and then you can apply a style to it to adjust it a bit.


This is Carnelia, btw.

Of course there’s also Artbreeder that everyone basically knows, but I don’t love it too much. Faces seem to be too similar to each other and as far as settings, it doesn’t respond to prompts the way I’d like.

From the ones I’ve tried, Night cafe is my fave so far because it offers a lot of flexibility, has a free tier and options to turn an image to high resolution once you generate something that’s worth the extra.
I might run a few abstracts created with Dream through Night cafe to increase their quality. I wonder how AI correcting work by another AI will come out.

What about you? Have you played with AI art yet? Which one?

Want to showcase your favorite creations?

UPDATE: Links!

If you give me other ai generator suggestions, I can add them to the list.

5 Likes

I like it, but don’t like the hoops I have to jump through to use much of it.

1 Like

I guess so but isn’t it the same when you commission someone to create art for you? You have to describe what you want and what style to aim for. Same thing with this. It’s not a stock photo, it has to be actually created.

The biggest difference is that in the end, it’s just a computer program and so it might misinterpret what you want. And there are limitations to what it can and can’t do.

But otherwise, it’s a really cool toy if you’re in the mood to play. If you’re not, that’s okay. More playtime for me.

1 Like

It’s not that I won’t at all, for example, I just tested this:

It’s the creating an account to do more than the 5 freebies it gives out.

1 Like

You get 5 credits every day and also a few extras here and there (for example, for liking # of other creations), but yes, you need an account.

1 Like

The poor thing didn’t really have a good focus with that spread, but I wasn’t expecting it to do much better.

I am cackling at this shirtless wonder it’s made this time.

2 Likes

My idea of what the Lust Circle of Hell would be like

"cyberpunk flying city
" - made with @NightCafeStudio

#aiart #nightcafe #digitalart

1 Like

Finished off the 5.

2 Likes

Oh, I tried Midjourney and already ran out of freebies.

I used it to figure out the general landscape of of my story.





I have lots of other locations/characters/creatures that I want to see, so I’ll def give the other AI engines a try.

I’ve also seen some people using edited/drawn over AI images as backgrounds for webcomics and other artworks and I really like that idea.

3 Likes

I wanted to see how close I could get to Yuki from School Live

4 Likes

Tapeworm waterfall?

I made this for @J.L.O euqjdKJerF4soeA0Gjkb–1–PBUK6|500x500

1 Like

I think it would really work for it and it would reduce the drawing time.

That’s what I just did yesterday, I modified the ai image in pixlr editor and it’s awesome.

A lot of this code is open source. There are instructions out there how to create your own generator. I think it would be really cool since you could train it yourself and maybe it would be more customizable, would bend more to your will. I haven’t tried that yet but I’m really curious what the learning curve on it is.

2 Likes

What’s that?

These pictures look super cool btw :heart_eyes:

2 Likes

Squid ocean.

2 Likes

It’s another generator. Just google it. You start the trial on Discord actually. There’s a limited number of images you’ll get for free so I suggest you try this process on a free generator first so you can get the hang of how the prompts work.

Tip: the first few words of the prompt seem to carry the most weight. Also, try modifiers to indicate style.

2 Likes

Cool! Thanks!

1 Like

I dunno what I’m doing wrong with dall-e, got the invite for the full site but so far they’ve been a let-down. This is the best thing they churned out for “a digital painting of an eagle with antlers”

(ETA) I mean, it’s not bad, but I expect to see the ENTIRE eagle with the ENTIRE antlers, not a cropped close-up that requires us to fill in the rest with our own imaginations.

the rest were barely more than generic eagle paintings you could easily find through a google image search.

You could try more words in the prompt like “standing on top of a tree” or “flying.”

Or try a different generator.

Also, remember that the first few words carry more weight. Digital painting of an eagle will give you different options than Eagle digital painting

Make something for me related around the churroverse you botlet.

2 Likes