Connecting through sports also capitalizes on one of the most culturally acceptable ways for guys to hang out.

We ran one hilarious, I mean, hilarious and sad study where we asked men and women who are fans, "How likely would you be to text someone you just met saying, 'I like you.

We should be friends.

Let's spend a few hours together this weekend?'" The other condition, "I like you.

Let's be friends.

Let's watch a game this weekend." Women equally, because, like, spend a few hours together, you could do lots of things.

Men, only watching a game.

He's seen that this technique works better than most people think.

Todd Rogers, a professor at Harvard, took that same core idea, social comparison, inspired by Opower and Bob, and applied it to a completely different problem, chronic absenteeism.

So when I was in DC I got to know the founders of Opower and started working with them and their team a little bit.

When I sh- returned to academia, I shifted my research program towards how do we mobilize families to support kids outside of school.

And the very first project I worked on was a complete rip-off of Opower, which was we sent parents rounds of mail showing a easy to understand horizontal bar graph showing how much energy they used and how...

Or, [laughs] no, I already got confused in telling it.

Showing how many days their kid had missed school and how many days their classmates had missed.

And we did this district-wide in Philadelphia, and it was 50 times more cost-effective than anything else that anyone had ever done to reduce absenteeism.

And Todd Rogers at Harvard couldn't resist mentioning Bob's communication abilities, too.

are some of the examples of entirely new software categories that couldn't have in your opinion couldn't have existed before ai right we've changed things but oh yeah yeah it's

uh they're they're not wrong you know it makes um development a lot faster you know with the with our team of Four tech people, we built what probably would have taken 40 just two or three years ago.

But that's not really the truly amazing stuff that we found.

We've had some interesting tech problems that just were not solvable.

three or four years ago.

The biggest one for us, when we were looking at this and trying to figure out how to build this into a product, my business partner, Trey, he's been modeling sports on the side for, you know, eight plus years.

And I said, the problem with this is there's no way to, yes, you have some generic modeling here, but you still have to kind of be a data scientist to pull this off.

You know what I mean? How do you get up and create it so guys like me can look at it? Well,

are some of the examples of entirely new software categories that couldn't have in your opinion couldn't have existed before ai right we've changed things but oh yeah yeah it's

uh they're they're not wrong you know it makes um development a lot faster you know with the with our team of Four tech people, we built what probably would have taken 40 just two or three years ago.

But that's not really the truly amazing stuff that we found.

We've had some interesting tech problems that just were not solvable.

three or four years ago.

The biggest one for us, when we were looking at this and trying to figure out how to build this into a product, my business partner, Trey, he's been modeling sports on the side for, you know, eight plus years.

And I said, the problem with this is there's no way to, yes, you have some generic modeling here, but you still have to kind of be a data scientist to pull this off.

You know what I mean? How do you get up and create it so guys like me can look at it? Well,