These kind of AI-supported conversations are possible, but both clinicians and patients have to play their part.

It's not like it's just gonna happen.

I would say I'd have to sensitize both parties to the possibilities of AI in their situation and how it could best be used.

And that means we'd have to now start changing how we educate.

And part of the education would have to be how you can use AI to its maximum ability.

[gentle music]

Where should we allocate resources? Which patients are most likely to run into trouble? There's one particularly promising opportunity.

As we're speaking, people are dying of sepsis worldwide, uh, because they're not, not recognized of sepsis.

Uh, I'm curious about the spread, so spreading to other countries and maybe what proved universal and what proved very context-dependent, having to be changed and adjusted for different places.

Well, actually, it's theory.

Uh, as I say, it's like gravity.

Apples fall to the ground in every country, so the SAFE country wo- SAFE program works in every country.

The biggest problem is time.

Do we have time in the day to talk about patient safety? And I'd argue, yes, you do have time every day to talk about patient safety, because you're saving time if you're talking about patient safety.

So I've done it in, um, in Australia, Argentina, Sudan, Mozambique.

So I wonder if you can comment a little bit on that through your work as the fellowship director for the, the Patient Safety Movement Foundation, and maybe at the, at the same time, you can tell me about your work with boards and how boards can enhance safety or not and, and how we kind of put that in the genome.