For most of my career, we have been trying to make computers disappear.

Microsoft Surface Table made information physical and social. Kinect removed the keyboard and mouse. HoloLens placed digital information directly into the world around us.
Different technologies. Same dream:

What if people could use computers without first having to become computer people?

Then AI arrived.

Finally, we had the intelligence layer we had been waiting for. Computers could understand language, images, documents, context and intent.

So naturally... we gave everyone a textbox.

Somehow, after decades of trying to move beyond keyboards and menus, our great leap forward became typing better prompts into a chat window.

This talk explores what comes after the textbox.

We will look at what earlier attempts at natural user interfaces taught us, and how today's multimodal AI, agents, voice, vision and context can change the way humans interact with computers.

Because the future is not about teaching everyone to become better prompt engineers. The future is about computers understanding what people are trying to achieve.

We will explore the opportunities, but also the uncomfortable questions. When an interface disappears, so do many of the signals users rely on to understand what a system is doing. A more natural computer interaction requires more trust, not less.

Drawing on a decade of Microsoft MVP work with Surface, Kinect and HoloLens, and the multimodal AI agents we are now building, this talk connects the history of natural computing with the next generation of AI.

The question is no longer:
"How do we learn to talk to computers?"
It is:
"How do we build computers that understand us?"

Can be both breakout and keynote. Opinionated, practical, and deliberately allergic to the idea that every AI problem is a chat problem. Live demo uses camera and microphone; a recorded fallback exists for unfriendly Jaarbeurs networks (I know the venue.... :-) )