Robotaxi is overrated today. Even if Tesla solves FSD tomorrow, the Tesla network will generate stupid low revenue. Waymo is already experiencing this in Phoenix in which each of their robotaxi is only generating 2 rides per WEEK. It's ahead of it's time and people are not yet ready to fully trust a computer driving them around. This can only happen if you get people's butt in seats, and have them experience complex scenarios without interventions for months or years before one can fully trust the system(or else you'll just think Waymo's 5 million intervention free miles came from a straight line and the minute you sit in the car it's going to hit an edge case and kill you). This is seriously how humans think about the unknown. You can tell because those who haven't experienced AP think it's dangerous and those who have can't live without it after they get used to it.
So the crazy twist is if enough people have experienced Tesla's FSD (because they have it) and went through intervention free miles for a year or two, these people will ONLY want to trust a Tesla's robotaxi and not look at the competitor's. When you are putting your life on the line, you go with the safest option..the one you are familiar with the most. So I really don't know how these other companies can generate enough income from robotaxies without that human trust connection and just need to take their words for it. They really need to ditch all those sensors and make cars cheap enough so people can grow along with autonomy to build that trust when the car is affordable to own.