Seeing the announcement of Gemini 3.5 Flash, I felt that AI model evolution has clearly moved to the next stage beyond the binary choice of "speed or intelligence." Particularly, the concrete progress in integrating it into corporate workflows is interesting, showing that it has entered the practical phase rather than being just a technology demo. Personally, I want to actually try how much it can support designers' work in the UI and graphic generation domain.