Daisy Mae (Flux Schnell)

Just going through and looking at images that seemed like they should have been better to see if running the same prompt through Flux Schnell gives better results.

Here's good old Daisy Mae. I think she does better in this new AI model at looking like a cartoon hillbilly than she did in all of the Stability-based images. Maybe about the same as in Dall-E 3.

The only downside? Well, in the one image I ran, the bottom ended up with a fake copyright notice to a fake name. AI apparently saw enough cartoons with copyright lines during its training that it thought it needed to add its own to a new image. Since that confuses some people, I thought it was better to just crop it off. I don't want to encourage angry anti-AI activists to say even more stupid things. (Saying this is probably safe. If they actually read things, they wouldn't be anti-AI in the first place.)

Here's the Dall-E version, for comparison: