It's from Jeff Oyster.

As a Berkshire and Tesla shareholder, I would like to hear your thoughts on the potential financial effects to GEICO, assuming Elon Musk delivers on his fully autonomous driving goal.

On Tesla's most recent earnings call, Elon said, if you've got at scale a statistically significant amount of data that shows conclusively that the autonomous car has, let's say, half the accident rate of a human-driven car, I think that's difficult to ignore.

Assuming Elon succeeds in reducing accidents by 50% versus human drivers, wouldn't auto insurance rates fall to reflect the reduced underwriting risk, thereby adversely impacting GEICO's revenues and float, and perhaps margins, too?

Well, yeah, if, well, let's just take the extreme example.

Let's say there are only gonna be three accidents, and the United States next year, for some crazy reason that anything that reduces accidents is going to reduce costs, but that's been harder to do than people have done before, but obviously, if it really happens, the figures will show it, and our data will show it, and prices will come down.

I wouldn't, there've been a lot of people talk about doing that in the past.

I mean, General Motors used to be very big in the insurance business, and when Uber first started, they used some firm, which now is, I think Agito confirmed, they're close to bankruptcy now, aren't they, because of taking things out at the wrong prices?

Is that true?

Yep, yep, yep.

Yeah, yeah.

Insurance always looks easier than it is, and it's so much fun, because you get the money at the start.

And then you find out whether you've done something stupid later on, but it's a very tempting business when somebody hands you money, and you hand them a little piece of paper, but really knowing whether you're, I mean, if accidents get reduced 50%, it's gonna be good for society, and it's gonna be bad for insurance companies' volume, but good for society is what we're looking for so far.

You might find kind of interesting, I mean, the number of people killed per 100 million passenger miles driven, I think it actually, when I was young, it was like 15, but even post-World War II, it only fell to like seven or thereabouts, and Ralph Nader probably has done more for the American consumer than just about anybody in history, because that seven or six has now come down to under two, and I don't think it would have come down that way without him.

There have been some kind of fluke figures of what people did during the pandemic, which are quite interesting, because they didn't drive, immediately they didn't drive as many miles, but they drove more dangerously, didn't they?

Is that right, Ajit?

Yeah, yeah, so the point I wanna make in terms of Tesla and the fact that they feel that because of their technology, the number of accidents do come down, and that is certainly provable, but I think what needs to be factored in as well is the repair cost of each one of these accidents has skyrocketed, so if you multiply the number of accidents times the cost of each accident, I'm not sure that total number has come down as much as Tesla would like us to believe.

Tesla has been toying with the idea of writing insurance directly or indirectly, and so far, it hasn't really sort of been much of a success.

Time will tell, but I think automation just shifts a lot of the expense from the operator to the equipment provider.