Uber may have to pay Waymo or redesign its self-driving software
engadget.comI have not completely understood why Uber is spending money researching self driving cars when they could buy them later from a mass producer with a competitive advantage in car manufacturing.
A self driving car manufacturer could refuse to sell to them and just enter the rides market. A self driving car manufacturer could sell to them at prices that keep almost all of the extra margin from using the technology over drivers for themselves.
My understanding is that Waymo is doing option one in Pheonix right now.
I have not completely understood why Uber is at all a reasonable investment nor why it ever received so much VC funding.
There is no business model. Uber doesn’t have anything unique, least of all self-driving tech. There won’t be any sort of sticky network effect since consumers can simply switch to a different ride hailing app with the lowest price. When (if?) AVs become reality, the car manufacturers themselves can become their own Ubers or AV-subscription service with no need for a middle man.
I'm surprised that you don't see the network effect of Uber/Lyft as being very sticky? It's not as simple as everyone just migrating to another app. It would be really hard to start a new ridesharing company from scratch because of a chicken-and-the-egg type of problem. New riders won't use the app if there are no drivers, and new drivers won't use it if there are no riders. Back when Uber and Lyft first came out, they had the luxury of being first which gave them more wiggle room; people were probably more willing to go out on a limb and use the app even if there was a sub-optimal amount of riders/drivers. But a brand new app wouldn't have that privilege. Would people be as patient as they were with Uber and Lyft? Or if it takes more than 5 seconds to find a ride would they just switch to Uber where they know they can have a ride as soon as they push the button.
One way the market could be more competitive is if there was some sort of protocol for ridesharing where you could access drivers from all the different services on a central platform. Then the services would have to compete to see who can offer the best prices and the best availability at certain times/locations. It would give room for new services to pop up because the transition between using Uber and whatever new service would be seamless. Or rather, there wouldn't really be a transition at all. Today I use Uber, tomorrow I use Lyft, the next day I use whatever, just based on what ride is closest/cheapest at the moment I want to use the app.
Especially considering the OEMs advantage when it comes to vehicle pricing, maintenance cost etc. Uber seems like an odd investment. Waymo seems to have the advantage when it comes to self-driving tech, although its currently only tested under ideal conditions. And they have partnerships with plenty of OEMs like Fiat Chrysler.
If AVs become a reality, cost and therefore the pricing for the consumer appears to be the only thing that will matter for the success of the app.
Interesting that the identified software isn't in the list of what Levandowski was accused of stealing.
The Levandowski case is focused on just nailing him on the things they're super confident about, so I don't think it's surprising that the software in this case is part of that list - not least because they had no way of knowing Levandowski stole it until this independent expert verified it.
As I remember it, they had pretty good logs. The indictment mentioned "over 14,000" files he downloaded.
I just had a thought, wasn't this accident after the Levandowski thing? I wonder if Uber pulled a bunch of code for fear of infringing on waymo and put in this crappy hacky code. They probably didn't want the Levandowski case to appear to slow down their progress.