Uber's valuation is based on a formula that looks something like (raw brand value in an autonomous car world) - (burn rate)^(number of years until autonomous cars are widespread).

The valuation can be quite high if you assume that the value in the exponent is low.

But I personally think that the "autonomous vehicle revolution" is not going to pay off as well for Uber as they had hoped. It's becoming more clear that the market for autonomous vehicles isn't going to look like the market for human-driven ones -- and vehicle manufacturers are going to be looking to build their own self-driving auto services rather than sell self-driving vehicles directly to consumers.

Uber's biggest innovation was their app. Even complex apps like Uber are not hard to clone if you have any sort of budget. There's an assumption a lot of investors make about their brand value: it assumes that another, already strong brand does not enter the same market. Seeing as Tesla has made no secret of their intentions to compete directly with Uber in the ride-hailing space, I don't think that's a fair assumption. Would you rather call an Uber or a Tesla at this point?

And what does buying a Tesla in that scenario even look like? Do you buy a car and rent it out a la Airbnb when you're not using it? How much money do you get for that, and how much of a cut does Tesla get since it's their algorithms actually driving the car?

Don't get me wrong; the brand value is actually really high because hundreds of millions of wealthy customers worldwide know Uber. Even if the company fails spectacularly, someone will buy them for a few billion in a fire sale just for the brand.