I agree with what you've written about robotaxis, and Uber/Lyft already put a ton of data analysis into ensuring they have capacity where it's needed. But I don't think apps are going anywhere anytime soon, or in decades for that matter, primarily because there are economic forces in play that make them desirable for the network owners.
Just to clarify, I'm not suggesting that the apps will or should ever go away, but rather that with sufficient volume both ways could become plausible options. If an available Waymo happens to be sitting there waiting for a passenger, I don't see why it shouldn't let me just tap my credit card on the handle or something and tell it where to drop me off. Of course, summoning one or tapping my phone would ideally work too.
> I don't see why it shouldn't let me just tap my credit card on the handle or something and tell it where to drop me off.
I imagine because a huge part of optimizing fleet availability and distribution is knowing where you want to go before deciding which vehicle you should travel in.
Ah, yeah, that's a good point. I can imagine some potential creative workarounds (e.g. having certain rides or types of rides involve transferring between two vehicles, possibly with multiple parties per vehicle like Uber Pool), but whether they'd actually be willing to support that is another matter entirely.
