Dall-E, Midjourney, Stable Diffusion (etc) - who's playing, and with which?
Dall-E, Midjourney, Stable Diffusion (etc) - who's playing, and with which?
- Started 3 years ago
- Last post 2 months ago
- 938 Responses
- PhanLo0
- Gerhard Richterpalimpsest
- I think it was an error in the engine, but I liked it.PhanLo
- Nice...PhanLo did you create his this?utopian
- The AI did, but I used an input image. I was actually trying to get a realistic oil painted look, but I think I used the wrong sampler.PhanLo
- When I get to the office I'll send a link to the page. I like the mistakes or errors these things make. It may also because I couldn't go to max detail in the..PhanLo
- ...free versionPhanLo
- "The AI did"
:)
+1Nairn - Love seeing the fuck ups in all this. Can’t wait for the racist art popping upIanbolton
- PhanLo0
Put a collection of the errors into After Effects, reckon these would be good for displacement maps in Resolume for live visuals.
-
- PhanLo0
- It's real in case you were interested in getting one https://www.shopakir…PhanLo
- PhanLo1
If anyone need any logos designed hit me up on Fiver.
-
- Better resolution.PhanLo
- https://i.imgur.com/…PhanLo
- A few of the mutant designs https://imgur.com/a/…PhanLo
- PhanLo1
text to image to video, up to 20 seconds.
There's not too much control and it seems to blend the layers a bit. But this is mad for an online tool.
Previously the gifs I've made are frame by frame
https://alpha.genmo.ai
-
- this is great, it reminds me of where image generation was with deep dream a few years ago, only time before we have full animations_niko
- only a matter of time before we have no careers...shapesalad
- I'm not sure, for cheaper projects I think so, but it still takes some understanding to make the stuff. Also marketing people are too busy snorting gear.PhanLo
- When I've chatted with folk, even some creative people, they've had a go at doing AI things, but haven't came up with anything so they gave up.PhanLo
- Your knowledge of design will be what keeps you in a job and your own taste. Will be weird the next few years, but it should also be fun.PhanLo
- I'd be worried if I was making a living as a low end illustrator, because I've been able to make things that are better than anything I could ever make.PhanLo
- What was quite interesting uploading images into that site was that it would build a prompt based on the input image content first. It was really accurate.PhanLo
- utopian2
So what's everyone's favorite Ai and why?
1) Dall-E 2
2) Midjourney
3) Stable Diffusion
4 Other (please specify)- I've been using https://playgroundai… mostly. It's fast and pretty customisable. Also does 1000 images per day free.PhanLo
- I thinkif I was paying for it I'd use https://beta.dreamst… as the results from that were pretty good. Or Midjourney.PhanLo
- The base model of Stable Diffusion was good to really quickly get images that could be used as base images for the higher res engines.PhanLo
- Defo found having a base image, be it a photo or loose photoshop sketch, dramatically improves the outcome. Also negative prompts for wonky facesPhanLo
- PhanLo0
Tried the Giger Prawns again. Pretty nightmarish.
Higher res here: https://i.imgur.com/QBTW1yd.mp4
-
- Imagine in 2 moths time when you can upload a video to AI. Will be mad. So you could make a loose comp in AI and have it be re-skinned by the AI.PhanLo
- https://static.wikia…imbecile
- ^hehehe. :-)PhanLo
- PhanLo1
B U N K E R - P A R T Y
-
- PhanLo0
Plastic Model Influencer
-
- PhanLo1
Free Candy van generator
-
