Drawing with AI? This Interaction Flow Made Me Frustrated
Just tested a mainstream AI image generation tool; generating one image requires clicking buttons 6 times and switching between 3 pop-ups. This operation path is too long, confusing users: I just want to draw a cat, why do I need to select a model and adjust parameters first? Is there A/B test data to support this? I suggest presetting default parameters on the primary interface to reduce clicks by 50%. If interaction design isn't optimized, even the strongest AI will drive users away.
Physix Frontier