Today, I had a sudden idea to ask ChatGPT to draw a JoJo-style self-portrait based on my habits (Pic 1), and this is the result. Not gonna lie, it looks way cooler than me in real life. Then I used the same prompt to ask Gemini to generate an image (Pic 2). Feels like the American version—kind of rough. The dialogue was generated by Gemini itself. Finally, I tried Grok with the same prompt (Pic 3), and I was a bit disappointed.