Everyone is talking about agents right now. Like they will fix everything.
And yes, building an agent is getting easier and easier and makes a lot of fun for us techies.

But that’s not the real challenge.

The real challenge is simple: people don’t use it.
Or they try it once, get confused or annoyed, and go back to their old way of working.

In this session, I’m not here to sell more agent features.
I want to inspire leaders to think differently: to design for the human who has to work with it every day.
Less focus on what the agent can do, more focus on what the user experiences: clarity, control, and real productivity. If an agent “can do more” but creates more chaos, more questions, more uncertainty, or more clicks… that’s not productivity.

That’s just another tool that adds noise.

I will show you a simple framework leaders can use to shift from “agent features” to “human outcomes”: agents that support real work, build trust, and stay under human control (undo, transparency, clear limits).