Dispatch via app is a solid improvement over the old phonecalls or street pickup, automatic credit card billing is an improvement over cash, and in cities with shady taxi drivers, GPS tracking is an improvement over potentially hacked taxi meters.
So there is an obviously valuable business model for "taxi dispatch app" here even without a moonshot - just perhaps smaller than Uber's dream of being the only provider of personal transportation in the world.
Uber would need to start understanding how ordering of goods works, returns, and service/quality complaints.
They could do it, but there are a whole lot of domain specific problems and approaches to resolve and it's not clear that's it's long-term viable.
It's also the easiest part to copy, most taxi companies have already. Now try asking they investors why the spent so many billions of dollars to break into the taxi business?
I disagree: they don't make cars. That's a huge impediment. The prime position company is surely Tesla. Or, really, any car manufacturer.
If you are retrofitting existing cars you are wasting a ton of money and getting a half baked solution to a dedicated provider.
Actually developing self-driving tech and then deploying it is almost entirely unrelated to the business that Uber has developed thus far. It's an enormous undertaking, and they were starting from almost zero.
A taxi company has well understood financials and valuation and would never be given a crack at VC Lotto.
In addition, the kind of quality control that Uber/Lyft exert over drivers almost certainly places the drivers in the position of being employees which makes things even more unprofitable.
Being a good taxi company in the 21st century will mean good at self-driving cars.
What is a startup? Poised for explosive growth. Are commodity businesses poised for that growth? No.
Not to mention that transportation-as-a-service would also be a lot more convenient than car ownership. No need to refuel, repair, or otherwise maintain your vehicle; just grab whichever one happens to be nearby. You can have as many or as few cars as you need at any particular moment, and get a car of any model or paint color you want at the push of a button.
Not relevant to my assertion. I said when they become the norm, as in, when they're no more expensive as other cars.
> ... transportation-as-a-service would also be a lot more convenient than car ownership.
This means I can share a car with several friends, neighbors, or room-mates with little fuss. The problem with sharing a car today is licensing, insurance, and the inconvenience of the car being stuck somewhere when in use. If it acted, instead, as your own personal driver it can handle multiple trips concurrently. Two people go to the movies, so they're committed for X hours, and in that interim instead of sitting in a parking lot the car can deal with other things.
Uber's business model arbitrages the difficulty and expense of owning a car vs. the expense of buying a cab ride. If the car becomes more cost effective because of personal ride-sharing their margin evaporates.
Today a decent car would cost me about $600/mo. all-in. If I could buy a 1/6th share in a car for $100/mo. and be able to use it in short stints with little contention that's going to be cheaper than Uber if I use it more than a handful of times.
My point is that self-driving cars will become the norm long before they're "no more expensive as other cars" precisely because of transportation-as-a-service companies like Uber. By the time we start getting anywhere near the point where self-driving cars are "no more expensive as other cars", car ownership will already have been largely supplanted by transportation-as-a-service.
> If I could buy a 1/6th share in a car for $100/mo. and be able to use it in short stints with little contention that's going to be cheaper than Uber if I use it more than a handful of times.
I'm not really sure what you're trying to say here. It sounds like you're describing transportation-as-a-service, but with the additional restriction of you only being able to use one particular car that you "own" 1/6th of instead of any car that happens to be nearby at the time when you request it.
That sounds like a needless restriction, and negates many of the advantages of transportation-as-a-service (no need to refuel, repair, or otherwise maintain your vehicle, the flexibility to obtain as many cars as you want at any particular moment, the ability to use a pickup truck one day and a minivan the next, etc).
Also, why would that be cheaper than Uber? I think a service like Uber is likely to be far more efficient with the usage of their self-driving fleet than you could ever be with any one, individual car, even if that car _is_ shared between multiple people.
If you're used to taking an Uber on occasion, taking an AutoUber isn't a big deal, just like trying a new coffee at Starbucks.
Even if GM is selling L5 autonomous cars to everybody in 2025, they're going to cost as much as a car does. Why buy it when you can pay by the ride? And Uber wants to be the name you think of when summoning a ride.