About this Session
Most AI tools are built around prompts — but intent is richer than text. It lives in a rough sketch, a tone of voice, the sweep of a gesture, or the careful collage of visual references. The question is: how do you design interfaces that actually capture it?
Most AI tools are built around prompts — but intent is richer than text. It lives in a rough sketch, a tone of voice, the sweep of a gesture, or the careful collage of visual references. The question is: how do you design interfaces that actually capture it?
Maria Borysova explores the emerging landscape of multimodal AI interaction — where designers move beyond the keyboard and start working with the full range of human expression. From sketch-to-render pipelines and image-to-image style transfers to voice tone, physical movement, and spatial input, this talk maps out what it means to build AI tools that understand how you mean something, not just what you typed.
Critically, Maria goes beyond the input side and into the output side: how do you design the controls, sliders, and interface elements that give people real agency over the result? Task and context awareness, manipulation handles, and the art of keeping humans meaningfully in the loop.
A talk for designers who want to build AI tools — not just use them.
