Artificial intelligence already decides what we see, hear, and buy, but the real question is: does it understand how we feel?

This talk will explore systems that interpret multimodal signals —tone of voice, micro-expressions, text, usage rhythm, and interaction patterns— to infer emotional states and adapt experiences in real time. We will review practical examples: interfaces that soften their tone when detecting frustration, assistants that adjust the pace of information delivery, interactive spaces that modulate light and sound based on audience feedback, and products that personalize messages, notifications, and micro-interactions to reduce friction and increase satisfaction.

We will also discuss how to integrate these capabilities into the design of brands, services, and physical–digital environments, from defining objectives to prototyping and measuring impact.

And then, the other side: what happens when an algorithm “believes” it understands us and gets it wrong? How far should we allow an interface to read us emotionally? We will talk about consent, privacy, bias, and transparency, and about how to design with AI without losing empathy, human judgment, or responsibility. A clear framework for innovating thoughtfully and carefully.