A Verge article [0] has some discussion on this with regard to a motorcyclist that follows this pattern, and FortNine also covers this in a characteristically well-made video [1].
There's also the problem of autonomous (electric) cars being too quiet when they first begin moving, but that's easy to solve.
So I'm not entirely sure about their safety being "proven", but that's largely because the problem domain is enormous and the data somewhat open to interpretation. Overall "proved" seems too certain a word as of now, without narrowly defining the scope.
[0] https://www.theverge.com/2022/7/27/23280461/tesla-autopilot-...
Not to discount the obvious potential safety issues, just noting that Tesla is taking a very different approach that is currently pretty far from being able to complete trips consistently, let alone be measured up against human drivers.
Certainly there are good reasons why SFFD is not keen on Waymo and Cruise.
Self-driving (and AI in general) has a tendency to go catastrophically wrong for impossible to determine reasons (imagine if each hallucination was a potential traffic accident), in a way that is hard to predict and is essentially open ended in terms of consequence. Humans do very stupid things much much less often than fairly stupid things - no such correlation is offered by AI.
I don't think that's true. I live in SF and am not aware of a single catastrophe owing to self-driving cars, let alone enough such catastrophes to form a statistical sample from which we could reliably infer tendencies.