Image generation is the least interesting and most concerning aspect of AI.
I see lots of useful tools made related to translation and code generation, but image generation seems to be primarily used to deceive, harass, or embelish to the point of questioning what the point of a photo is anymore anyway?
I can give one innocent example - I made an app that summarises D&D sessions, including images to depict some of the scenes.
But I understand the jaded perspective. I guess the idea of ‘deepfakes’ didn’t give AI image gen a great reputational start.