The funniest part is I've asked Copilot, Gemini, and ChatGPT if they are aware that they suck at creating text in generated images and they are well aware. Gemini straight up told me that while it's aware that it needs to combine the text generation section of it's programming with the image generation section of it's programming, it doesn't have coding to let it effectively combine the two. I love making "AI" generate pictures with text in it. Always cracks me up.ah yes, the "SO" coyote, with "COΓCɿV" and "/ON ƎOƝ" written on the intake runners. a classic, really