Blueberryverse
patreon
Increasing Content Variety and Production (Important!)
🕑 Added 2024-03-12 21:27:50 +0000 UTCHey, everyone! Yesterday I started messing around with AI generation and Stable Diffusion for the first time and I've decided I'm going to incorporate it into my content creation process.
But fear not. The AI generated content is going to be an addition, instead of a replacement for the actual artworks I make. Think of it as filler content for when I'm too busy to upload regularly. I'll even have more time to focus on Gumroad products without having to worry too much about my Patreon going silent.
How is it gonna work?
1) Here is the workflow I've adopted for now. First, I come up with a very basic sketch of what I want the result to look like.
1.1
1.2
1.3
2) After combining it with some personal prompts and fine tuned settings, the AI gives me a base I can work on.
2.1
2.2
2.3
2.4
3) Then I take the output straight from Stable Diffusion to Photoshop and I start making adjustments, changing up stuff, eliminating mistakes and all. Basically making it look like what I had in mind originally.
3.1
3.2
3.3
3.4
All the AI generated content is going to be uploaded to a specific collection, so it will be very clear which works are made with AI and which ones are not. All tiers will have access to it.
Also: During the next days I'll be working on my own personal model, trained with my 3D renders. Wish me luck!
Comments
Blueberryverse
Nah, it's actually pretty easy to set up. Some people do it through Google Collabs, but you can also do it locally. Since my rig meets the minimum Vram requirements, I've decided to run it locally. From there it was just a matter of experimenting and learning from trial and error. It's not as hard as some AI bros claim it to be, though. I went from making absolute aberrations to the results I've shown above in a single day.
Cyber5la5her
How does one even use stable diffusion, don't you have to like program the AI?
