The whole article basically says that Travis says that he'll buy half a million Tesla cars in 2020 if Tesla can actually make the cars fully autonomous by then.
Fully driverless cars are still not going to be a good thing for Uber.
1. Uber currently has the ability to grow and move at software speed. Other people buy the cars that it uses and assume the risk. If it needs to pull out of a city because of hostile regulation or whatever, its drivers are left holding the bag. Going from that to being a company with an expensive and gigantic physical inventory would be incredibly painful for it.
2. Driverless cars are the end-point of the commoditization of the rides-for-hire market. If you're buying a Tesla that drives itself, anyone else who provides the same Tesla provides exactly the same level of service to your customer. You can only compete on price or availability -- but availability now means buying expensive physical inventory, so both means of competing are now super, super expensive for Uber.
3. All of this assumes that it even makes sense to have a rides-for-hire business in a driverless car world. It's possible it will, but we don't know that! If the sensor package of driverless cars is relatively cheap and relatively durable in value, then probably driverless cars will actually drive business away from rides-for-hire and towards car ownership.
1. The recent California ruling might minimize this. If drivers are employees, it's not nearly as easy to leave them holding the bag. The physical inventory of cars is also a competitive advantage. If you want ride that's great, because I happen to have 5 cars in your neighborhood right now...
2. That's no different then the current position. I can use Uber, Lyft, a taxi or buy a car myself. And like now, the competition is on who has a car close to me and who has the cleanest car. I could personally compete with Uber by have one well maintained clean car with a mint on the dash, but I could only afford one.
3. Sorry, but I think you're arguing against #1 here. Uber having a large fleet of cars with the logistics to handle them would kill my small business's desire to own cars. The goal with driverless cars is to keep utilization as high as possible; if Uber can keep their cars at 95% utilized they will probably be cheaper than my car which is about 5% utilized.
Your point 1 is "Maybe Uber is going to get a huge kick in the teeth right now, so then the later kick in the teeth will... pale in comparison?" But actually they aren't terribly related. You get why startups exist in the computer sector and not in the traditional manufacturing sector? It's because growth is so much easier in the computer sector. The marginal cost to add a customer in software is tiny. The marginal cost to add a customer in a business with physical inventory is much higher. Going from software to physical business is not to be desired.
Your point 2 is yet more awful for Uber: Great, they get to compete on cleanest car? So now in addition to having to buy these very expensive depreciating assets, they also have to add local employees everywhere to actually clean them? Wasn't the advantage here going to be that they took the cost of "paying people" out of the equation? Uber's growth is going to be based on their ability to build out a really competitive cleaning service? Does that sound like a business you want to invest in?
Point 3 is interesting:
So, look, people who were already of the opinion that there's too much car ownership have gotten excited about the potential of driverless cars to increase car utilization. But "increasing car utilization" isn't a goal of normal people. "Saving money and/or receiving more value" is the goal.
A heavily utilized Uber car is cheaper on some margins than a lightly utilized personal car (that is, it's cheaper based on time-based depreciation and on parking fees). It's more expensive on other margins (it drives more miles per passenger served than the personal car, so its fuel costs, maintenance costs, insurance costs, and milage-based depreciation are higher). And of course Uber has to make a profit, so that's an additional cost as well.
How does that all play out? You can add up the hypothetical costs fifteen different ways based on different assumptions, and come up with whatever numbers suit you. The truth is that nobody knows how the cost of ownership of driverless cars is going to break down yet.
But I will tell you one thing: No car, Uber or otherwise, is going to be 95% utilized.
Because there is no reason for the manufacturer to sell to Uber and Uber has no value proposition if driverless cars become real. If you are the builder of driverless cars why would you want Uber to dominate the market and be in a position to eventually dictate terms to you when you can instead sell at a premium to Uber's desperate competitors or just create your own car agency and laugh all the way to the bank as Uber becomes the next WebVan...
They don't have exclusive access to cars currently and that doesn't seem to be a problem. There are a lot of people who won't buy cars if autonomous cars are available.
I guess I just don't see there's much of a barrier to entry now, so autonomous cars aren't much of a negative.
Fully driverless cars are still not going to be a good thing for Uber.