2025-03-27 After the ChatGPT image generation update on the 25th, my feeds on X and Threads were suddenly filled with people transforming their photos into Ghibli-style, Doraemon-style, or dramatic manga-style images. I was surprised when I prompted the model to redraw a certain illustrator’s work in a Doraemon style—and it came out looking just like Nobita. It made me reflect on how we perceive and interpret what we see. Being able to describe what you like is closely tied to how you create prompts or tags. But that doesn’t always have to be in words—it could just as well be a sketch or a photo.Maybe what matters is how we treat the medium itself, or even the method of communication, as a kind of brush for expressing intent.