Imagine what will happen when we combine decentralized car use with autonomous driving. Uber is already cheaper than driving yourself around, but think how affordable it would be if no Uber driver was needed at all.
Uber is only cheaper than driving yourself around if you points of comparison involve low-mileage use cases, a high-density area, an expensive car, and expensive gas. In other words, a wealthy individual who lives in San Francisco, New York, Chicago, or LA.
For everyone else, Uber is far more expensive than owning their own car.
LA isn't a high-density area - it's very spread out, with lots of low-medium density housing.
How is Uber doing in LA, I wonder? Just watching it here on HN, you could be forgiven for thinking it was SF-only (due to the demographics of the site)
The LA metropolitan area may not have as high a density as SF, NY, or Chicago, but that's primarily because it's several times the geographic size of those metropolitan areas and includes lots of uninhabited space. Many parts of LA are just as dense as similar areas of SF, NY, or Chicago.
Te type of customers who would use Uber in LA live in the high-density areas (generally K-town, Hollywood, West LA, and DTLA), and Uber does well in those areas. For anyone living in the low-medium density areas, i.e., the suburbs, Uber simply isn't economical.
I lived in Santa Monica, which wasn't too high density, but I recall that taxi prices were absolutely outrageous. Hopefully Uber has caused those to drop some.
This is what excited me about Google buying part of Uber -- self-driving Uber taxis. Combine that with electric vehicles that can intelligently remove themselves from the swarm of cars and be replaced by a charged vehicle, and you have on-demand, cheap, and potentially environmentally more sound taxis.
Why do you think the consumer will see any significant price benefits from automation instead of having the profits diverted to shareholders, executives or as reinvestments?
Google isn't the only company developing autonomous cars. Multiple car companies have own working prototypes. So it doesn't look like there's any monopolization of the technology itself. It seems likely that the capital cost of an autonomous car will be roughly comparable to the capital cost of a normal car.
Sure, but a self-driving car by itself is of fairly limited utility in a taxi operation; the advantage that they provide comes with the data gathering and analysis system that provides centralized control of the fleet. Automation, as it matures, should drive down the per-car cost and drive up the competitive advantage of the player that can expend the most on central resources.
When you reduce the relative importance of the per-unit cost and drive up the relative importance of centralized costs, you make a market that has stronger natural tendency to form a monopoly, because there is more of a positive feedback loop reinforcing the position of the leading player.
