[1] https://www.reinsurancene.ws/waymo-shows-90-fewer-claims-tha...
[2] https://pmc.ncbi.nlm.nih.gov/articles/PMC11305169/
[3] https://pubmed.ncbi.nlm.nih.gov/39485678/
[4] https://www.researchgate.net/figure/Comparison-of-Swiss-Re-h...
[1] https://www.reinsurancene.ws/waymo-shows-90-fewer-claims-tha...
[2] https://pmc.ncbi.nlm.nih.gov/articles/PMC11305169/
[3] https://pubmed.ncbi.nlm.nih.gov/39485678/
[4] https://www.researchgate.net/figure/Comparison-of-Swiss-Re-h...
I hear people on this site complain about these things all the time.
It's important to realize the reason for that.
Crashes by human drivers are hugely disproportionately by people who are driving drunk or with insufficient sleep or significant distractions etc. In other words, it's not a difference in the cars, it's a difference in the drivers. Waymo can beat a drunk driver, and therefore can beat the human driver arithmetic mean which has the drunk drivers averaged in.
That doesn't mean it's any safer than driving an ordinary car when you're not drunk.
The computerization of formerly mechanical features making it harder to DIY repair is a separate but also valid concern, though I'm not sure how it applies to EVs.
Added: see https://x.com/IntCyberDigest/status/2011758140510142890 for exactly the kind of thing that nobody wants.