> If Tesla's APIs can't handle input spam, that's on them. That's a very predictable and easily-solved problem. The entire point of an API is to expose a limited surface area so that you can easily interoperate while maintaining control and responsibility.
Sure, it's on them, and it appears they've solved that problem by cutting out other apps entirely. Certainly I'd prefer it if they would let people experiment with their product, but I can see why they wouldn't want them to, considering that it's a rapidly evolving, high-stakes product.
> If Uber sends your car to the wrong location, then that's that. No big deal. People go to the wrong locations themselves all the time. I don't see how this is an Uber-specific problem, and I don't buy the argument that people are smart enough to hook their Tesla up to Uber but too stupid to understand who's at fault when the car correctly navigates itself to an incorrect location.
It's not "Uber-specific", it's just an example of the sort of thing that could happen if a third-party app makes errors. Reasonable people will understand that it's the app at fault, but most people will just see the headline: "Panicked partiers trapped as Tesla self-driving car goes on a robot joyride!"
> There is no "approved Tesla tool" to connect to Uber. What happens if they modify their "approved Tesla tools" so they won't auto-drive you to a Burger King?
Tesla disallowing third-party apps and self-driving-for-revenue outside Tesla Network doesn't make a difference to whether they could do such a thing. But you can see that they have sensible reasons to not want these things (software concerns, commercial insurance, public/regulator perception) and there's no real reason that they would block you from going to particular destinations. Internet Explorer could block you from using anything but Bing, but they don't. That would be ridiculous.
