I suspect that “Stable Diffusion” and Midjourney are already being used more than Dall-E 2 (and that the Chinese and Russian Dall-E counterparts have limited short-term impact since there are many technical models of AI already).
Earlier I had provided news about MidJourney and the open source Stable Diffusion, but unfortunately this was not picked up by the Tweakers. I think those two are now the size of a Dall-E 2. (and perhaps more important). Mobile apps, including:
– https://beta.dreamstudio.ai/ (From the developers of Stable Diffusion)
– https://pixelz.ai/
– https://www.starryai.com/
– https://nightcafe.studio/
– https://www.artbreeder.com/beta
– https://replicate.com/ (Introduce a lot of AI “models” to use/upgrade/transfer the pattern)
– https://huggingface.co/spaces/stabilityai/stable-diffusion
– iOS app “Wonder- AI Art generator”
Photoshop add-on: https://www.getalpaca.io/ (in private trial)
Photoshop add-on: https://christiancantrell.com/#ai-ml
It also appears that this version was made a year ago.
Nice too, op https://lexica.art/ You can find almost all the images created in the Stability Diffusion beta beta on Discord. I can also find my photos there .
By the way, the impact of AI art extends on several fronts:
– Inspiration / Moodboarding for illustrators: https://youtu.be/GwagpfL_BEU
Make videos with music: https: //www.instagram.com…M-cX/? igshid = YmMyMTA2M2Y =
For backgrounds in web designs: https://youtu.be/-Emf7un0Wt4
Create artistic assets for games: https://www.youtube.com/watch?v=blXnuyVgA_Y
It is clear that the big three of Western AI (Dall-E 2, MidJourney and Stable Diffusion) have their own “style”. They can also complement each other. You can use the output of one as the input to the other and then request variations or improvements, change parts with interior paint, or enlarge images with exterior paint.
It doesn’t end there either. You can roll out your output through all kinds of AI, for example AI that specializes in repairing and optimizing faces and then through AI that does upgrade. There are already tools like chaiNNer that allow you to run an image across a series of AI systems.
If you want to learn more about the art of artificial intelligence, these are currently good resources:
– “MattVidPro AI” YouTube channel – https://www.YouTube.com/channel/UC5Wz4fFacYuON6IKbhSa7Zw
– Scott Dettweiler’s YouTube channel – https://www.youtube.com/c/ScottDetweiler
And some of the beautiful pictures I’ve created that are on my Instagram:
Honorable Owl https://www.instagram.com/p/ChNwPtJjDy-/ (stable spread)
– Cat companions roaming – https://www.instagram.com/p/ChVWYOdjnIG/ (stable spread)
– A beagle cat with a hat – https://www.instagram.com/p/Cgk8DsLDIkz/ (Midjourney)
– Cars and lightning – https://www.instagram.com/p/ChQWGJEjy7A/ (Midjourney)
– Some cars with fog and light effects – https://www.instagram.com/p/ChQVQwEDD42/ (Midjourney)
– Snow fox – https://www.instagram.com/p/ChYIH6MDHnj/ (stable spread)
I spent a few hours on most of the images by adjusting the prompts and displaying the differences. But I couldn’t draw them myself .
[Reactie gewijzigd door spellcoder op 15 september 2022 13:40]