I'm not going to pretend to understand the intricacies of Lyft and Uber and their experiments with food delivery, carpooling, and other endeavors. But scaling up does not solve the problem of paying your contractors more than they're making you - in fact it only scales up the problem.
Granted, I've thoroughly enjoyed my cheap rides subsidized by VCs for years. But it's just not a viable business, and I have never heard a convincing argument to the contrary. I feel as though few VCs are looking to create value so much as they're looking to get in early and cash out before the house of cards falls over.
If they had to worry about owning and maintaining all their cars, suddenly they're in a new and very capital-intensive line of business. I'd expect at least one of the established rental car chains- which have physical infrastructure like motor pools, mechanics and relationships with local auto shops, and logistics experience with all of the above- to eat Uber's lunch!
Every way I look at it, work on self-driving technology by rideshare companies is 100% smoke and mirrors to drive investment, and not a serious business plan.
Of course, that depends on a large barrier to entry, and I can't see how the barrier to entry on that market could be large. But well, VCs clearly disagree with me.
Also I cannot imagine Uber to be in any different situation.