>Could it be that Waymo should just be able to stop the car?
That kind of remote control opens up the possibility - maliciously or accidentally, likely or unlikely - for every Waymo in the fleet to abruptly stop, regardless of whether it's safe to do so. That scenario is orders of magnitude worse than "sometimes a Waymo gets lost in a parking lot and it takes a thirty second call to fix it".
You don't think Google is able to stop a Waymo vehicle? They can literally control it remotely. The issue is that the person behind the phone did not have the authorization or access.
Of course not, but I'm open to the idea that they should be able to escalate to someone who can. This is a transitional period with self-driving cars, and it helps to mitigate serious potential safety issues.
Ideally, in the future you would have to expressly press some button to send an OTP to someone you wanted to allow to control your car, with fine-grained permissions.
Will that future happen? Probably not, and we'll probably see extreme corpgov oversight and mass reduction in individual freedom in return for convenience.
Anyway, I was only responding to OP's claim that support for remote access capabilities could lead to an exploit: There's already remote access capabilities in these cars.
That kind of remote control opens up the possibility - maliciously or accidentally, likely or unlikely - for every Waymo in the fleet to abruptly stop, regardless of whether it's safe to do so. That scenario is orders of magnitude worse than "sometimes a Waymo gets lost in a parking lot and it takes a thirty second call to fix it".