3 ms·

Uber had better hope that the startup costs of self-driving vehicles are a lot lower than $36B. They can't afford that.

brudgers 9y ago
Uber spent $680 million on the Otto. GM spent $1000 million on Cruise Automation. Neither of these came with tooling and supply chains or production engineering. In the automotive industry, a billion gets some talent and potentially promising intellectual property, not vehicles on the road. Yes, I think the Otto purchase was a bigger waste of money than the subprime leases. On the other hand, it didn't wreck lives by treating people disposabley.
- aetherson 9y ago
  I mean, sure, but on the other hand $1B is 36 times lower than $36B. Maybe we're just talking in cross purposes. Obviously if you buy a million cars, you'll be spending tens of billions of dollars. I didn't mean that Uber was hoping to be able to roll out a massive worldwide fleet of driverless vehicles for under $36B. I meant that Uber has to demonstrate a mature technology and business model involving lets say thousands or tens of thousands of vehicles that people are actually paying for and that have unique advantages over their many competitors, for much less than $36B. If they can do that, they won't have any difficulty securing the financing to scale their fleet of vehicles up significantly. If they can't do that, and they can't do that for much, much, much less than tens of billions of dollars, then they're in a lot of trouble. I do think that we tend to get sloppy about large numbers. It feels like people hardly see any difference between $1B and $5B.
  - brudgers 9y ago
    Cars that are produced in the 1000's or 10,000's tend to run in the $100,000's [1] even when the costs are amortized over the total production of an established manufacturer where startup costs like site acquisition and robots and hiring are already sunk costs. To better explain, the difference between one and five billion is significant when its my money. But it is not a quantum difference in terms of ordinary automotive manufacturing. Tesla is fourteen years old, has raised billions of dollars and is not relevant at the global scale in terms of production or sales. [2] More importantly, Tesla's differentiation is not self driving technology, it is electric drive trains which have an automotive history going back to the 19th century. Tesla is able to draw on established power technologies and engineering expertise in the production of batteries and electric power systems. Just maybe it will be able to produce a few tens of thousands of $35,000 electric cars a year at some point. I think that the odds Tesla will be first with a full on self-driving vehicle with wide availability are pretty low...but given that it has been working on the idea for about a decade or so, much higher than Uber's ever were. [1]: To be clear, I am not talking about 2017 Camero's with green leather interior and cherry red paint. [2]: That's not to say it does not have potential to become so over the next fifteen years.
    - aetherson 9y ago
      Uber doesn't have to manufacture cars from the ground up, and has given no sign that I'm aware of that they're interested in doing so. They'll use after-market additions of their self-driving technology, and most of the manufacturing will be done by third parties, for a long time yet. If they get to the point where they're even thinking about creating automobiles from the ground up, it'll be past a bunch of hurdles. Even then, they'd still almost certainly partner with an existing auto-maker.

Uber spent $680 million on the Otto. GM spent $1000 million on Cruise Automation. Neither of these came with tooling and supply chains or production engineering. In the automotive industry, a billion gets some talent and potentially promising intellectual property, not vehicles on the road. Yes, I think the Otto purchase was a bigger waste of money than the subprime leases. On the other hand, it didn't wreck lives by treating people disposabley.

aetherson 9y ago
I mean, sure, but on the other hand $1B is 36 times lower than $36B. Maybe we're just talking in cross purposes. Obviously if you buy a million cars, you'll be spending tens of billions of dollars. I didn't mean that Uber was hoping to be able to roll out a massive worldwide fleet of driverless vehicles for under $36B. I meant that Uber has to demonstrate a mature technology and business model involving lets say thousands or tens of thousands of vehicles that people are actually paying for and that have unique advantages over their many competitors, for much less than $36B. If they can do that, they won't have any difficulty securing the financing to scale their fleet of vehicles up significantly. If they can't do that, and they can't do that for much, much, much less than tens of billions of dollars, then they're in a lot of trouble. I do think that we tend to get sloppy about large numbers. It feels like people hardly see any difference between $1B and $5B.
- brudgers 9y ago
  Cars that are produced in the 1000's or 10,000's tend to run in the $100,000's [1] even when the costs are amortized over the total production of an established manufacturer where startup costs like site acquisition and robots and hiring are already sunk costs. To better explain, the difference between one and five billion is significant when its my money. But it is not a quantum difference in terms of ordinary automotive manufacturing. Tesla is fourteen years old, has raised billions of dollars and is not relevant at the global scale in terms of production or sales. [2] More importantly, Tesla's differentiation is not self driving technology, it is electric drive trains which have an automotive history going back to the 19th century. Tesla is able to draw on established power technologies and engineering expertise in the production of batteries and electric power systems. Just maybe it will be able to produce a few tens of thousands of $35,000 electric cars a year at some point. I think that the odds Tesla will be first with a full on self-driving vehicle with wide availability are pretty low...but given that it has been working on the idea for about a decade or so, much higher than Uber's ever were. [1]: To be clear, I am not talking about 2017 Camero's with green leather interior and cherry red paint. [2]: That's not to say it does not have potential to become so over the next fifteen years.
  - aetherson 9y ago
    Uber doesn't have to manufacture cars from the ground up, and has given no sign that I'm aware of that they're interested in doing so. They'll use after-market additions of their self-driving technology, and most of the manufacturing will be done by third parties, for a long time yet. If they get to the point where they're even thinking about creating automobiles from the ground up, it'll be past a bunch of hurdles. Even then, they'd still almost certainly partner with an existing auto-maker.