Slightly tangential, but I hadn't seen the Tesla article before. It's really interesting that they're bleeding like that, and I can't help but wonder if it's because of the management style.
In EVs and rockets, Tesla and SpaceX are the place to be. They're new, exciting, fast paced and nobody else is pushing the field forward, so engineers will put up with a lot of crap to be involved. But self driving cars? You have TONNES of options, and Tesla isn't really the place to be, autopilot notwithstanding. Google was one of the first heavyweights in the space, and now everybody is gathering teams. Why put up with the Tesla stress?
As for the taxi service, my bet is the last 1% or 0.1% of test cases is being an absolute bear to try and fix. And that's the real gotcha of self-driving cars, right? Ironing out the minor edge cases.
