[lvl155]

Aren’t you PE? Just look up the lawsuits bro. On a serious note, I don’t think Google or Tesla will take on that liability once we get to scale. That basically defeats the purpose of autonomous. Their legal and ops team will do everything to push an alternate business model similar to lease. This is why I really think autonomous has to be at least two orders of magnitude “safer” to be viable at scale (more than 10% on the road).

[Dylan16807]

Comparing to taxi rates and positing a car that's pretty good by human terms, $500 a month for insurance defeats the purpose of autonomous? It doesn't seem like a big issue to me. That's less than a dollar per ride at unimpressive safety levels, so I can't imagine why it would need to be 100x safer.

[lvl155]

Because you’re going to amplify “errors” and “failures” if you have a system of cars running same or similar autonomous models.

[Dylan16807]

Amplified meaning if you have 1000x as many cars you have 1000x as many failures?

That's fine. Insurance is built on that model. What's the issue?

Or do you mean something else?

(If amplify means a ton of the cars crash on the same day because they share code, I doubt that being a big effect, because all the cars are in different places working on different data feeds. And even moderately high spikes would be absorbed fine.)

(If amplify means the crash rate of each car goes up when you add more cars to the fleet... why would that happen?)

[JumpCrisscross]

> Aren’t you PE?

No. VC.

> Just look up the lawsuits bro

Can you name one? I’ve been in a single taxi accident. Liability was never even questionably mine.

> don’t think Google or Tesla will take on that liability once we get to scale

Based on what? Centralising liability tends to facilitate its transfer.