Agreed for their current model. However Uber is very different to Visa.
What's the endgame for Uber? Surely that's operating an autonomous fleet. No drivers. Visa will always have merchants, but Uber will not always have drivers.
What's the endgame for Uber? Surely that's operating an autonomous fleet. No drivers. Visa will always have merchants, but Uber will not always have drivers.
Right now, Uber's core competencies are having a nice app, and managing a bunch of people-centric things like recruiting drivers and getting them to be available to drive at times when people want rides. Of those two, the latter is probably the more difficult one to clone. It's also useless to a company that doesn't employ any drivers.
What they have close to zero competency in is fleet management. That's something for which traditional cab companies have something approaching a 100 year head start, and they which are very good at.
I think that's the reason why Uber has decided that it's so critical that they develop their own self-driving car technology: They need to control a supply of self-driving cars during the critical transition period if they want to avoid being at a disadvantage while it's happening. If everyone's buying commodity self-driving cars from a big automaker, then it's just a game of who can operate most efficiently in order to offer the lowest prices. Traditional cab companies were already good at that, and, if anything, the influence of Uber on their market over the past however many years has forced them to get even better at it.
There is no reason why municipalities should not consider their own autonomous fleets as part of public transit. The roads are already public, having a fully autonomous fleet is hardly any different than public buses or trains purchased from private manufacturers. It'll be a commodity (or at least it should be, short of monopolization and cronyism).
Except that they're terrible inefficient, and still 10~15 years away. Even if the tech was available tomorrow, and even if you could cut off Interstates so they only allowed autonomous vehicles and all those vehicles could run at 120kph, bumper-to-bumper at less than .5 meter intervals, a six lane highway would even begin to approach the capacity of a single track of London train line running at <2min intervals (which they often do during the morning rush .. often times < 30sec intervals).
http://penguindreams.org/blog/self-driving-cars-will-not-sol...
Self-driving cars are new and shiny and all the rage, but they're nowhere near ready for production deployment en-mass, people still don't realize that no person can actually own a autonomous car (the inspection/maintenance cycle would need to be more strict than a small engine aircraft) and therefore they wouldn't be able to be modified in any way.
For a fraction of the cost, American cities could revitalize and expand our crumbling rail infrastructure, which would be way more helpful than autonomous public car fleets.
The point is car ownership itself is obsolete and it's not a big stretch for there to be a public car option. The road infra is already public, just like rail tracks and rail cars.
As far as autonomous vs traditional public transit, it's a false dichotomy. Municipalities will use both when it comes to moving people around, because they're good for different use cases. You choose one particular metric where trains do win out, but that's an irrelevancy, because it's not even close to feasible to build a system of trains that do everything that cars currently are able to do.
All those cost points don't benefit the endgame.
If Uber’s big moat is on the driver supply side, an autonomous fleet could wipe that away and hand the profits to Google & Apple in an auction based market for rides.