Tell us a little bit of the origin story and why it is that's your name.

So three startups ago, I started a company.

I learned about this technology called machine learning.

It was a supply chain startup, and we're trying to pull data together from multiple companies and recognize from disparate data from different companies that bought from the same suppliers, patterns of data that we could actually, instead of having humans understand what was going on, can we actually have computers run through the data and pull patterns out that help companies understand? supplier performance, past performance, and predict future performance.

And there's a number of things that I learned from that.

But it wasn't cool enough to call it artificial intelligence back then because it had just been sort of discovered.

And the generic term of what was going on was the machines learned from the data.

We couldn't get there.

How serious is that pressure now as the hot weather continues and the, and the drought goes on?

It's really serious.

I mean, of course wildlife needs, you know, exactly what we need.

It needs water, it needs food, and it needs shelter.

And, you know, the longer the drought goes on, the more the pressure is put on, on all three.

Of course as plants start to shrivel up and die, we see less shade for wildlife, less spaces for animals to call their home.

We see ponds starting to dry up, which, you know, can affect all sorts of things from our dragonflies, you know, that need water to breed in, our frogs and things that need to live in these wet habitats.

And in the longer term, do we need to be seriously looking at how we manage land?

So could you maybe comment on that and the implications of this regulatory approach and the importance of explainability and how you guys are addressing that?

Well, we see explainability from two perspectives.

One, regulatory, and I'll get to that in a second.

But when users try to look at this as what can it do for me or to me or does it help me or hurt me, explainability is sort of that last mile.

If the models are perfect, and they never will be, but if they were, There's still a lack of confidence in a number of people when this is new to them from a standpoint of can I trust it and can I use it? And if it gives them the right number or the right dollar amount for an expected, how to price a policy or how much a claim is going to cost or how long it's going to remain open or many, many other potential answers.

If the person is not confident as to how the model works, came to that conclusion, if they don't agree or they're not certain, they can, in fact, just pass on, even though the model might have given them some important directional information.

And so the explainability allows the person to say, OK, I see how the model got there.

and then obviously machine learning has been around for a long time now there's llms there's slms uh gradient in its form is gradient i think it's been around since 2018 it was within milliman before that that's correct curious about sort of what the evolution of ai has been within the business kind of where it is today where you maybe see it going in the future

Mm

... and, and people didn't wanna leave, you know, uh, home- ... for Christmas.

Although, I think players are doing that now 'cause they've gotta get here in time to be ready for this, the tournaments, uh, starting.

So, um, I don't know the answer to that one.

Mm

... from yourself, what you expect from, uh, the media, from your friends and family, and- and- and handling all that situation is, uh, certainly when you're in the final, it's, uh, it's very special.

So if you haven't experienced it yet, once experienced it-

I know that we're kind of moving towards that curve of adoption, but could you maybe speak to a little bit more about model accuracy in general, but then more specifically gradient models?

I think the black box has been the term many have used for a long time.

I think actuarial math that has really dominated insurance is more about knowing what math was used than actually trying to get an accurate outcome.

So it's been about transparency.

So the regulator could see that you use the proper method to project your losses so that I can sign off on your loss projections and things like that.

There are reasons for this.

But when you get into other things like pricing risks or estimating the duration and or complexity of a claim, some of that math, that's not the best math for those different applications.

And how do you sort of build a moat then to kind of keep ahead of the curve and the rising tide of AI?

And we wondered how that, or first of all, what did that mean to be an Australian Open champion? But second of all, did that double success pave the way for your single success in the years that followed?

Yeah, I did.

I started out really doing well in doubles with Bob in the late 60s, early 70s.

We won a few doubles titles.

We won the U.S. Open.

We won here.

We lost the finals of the French a couple times.

Tell us about the inspiration behind writing it and I guess what kind of like lessons or values or philosophies you're pouring into it that you would like people to take out of it.

Is he still there?

I shot him.

I think I killed him.

I think he killed my (censored) wife.

I couldn't (censored) protect my wife.

They're both (censored) dead.

We have a law in place, you have a right to, to use deadly force if you think your life is in, in danger.

But to police, Bryan Capnahurst was starting to look more like a victim than a killer, and this was now a different kind of case entirely.