Of the various promises Elon Musk made while discussingTesla's (TSLA) - Get Report second "Master Plan," none raised more eyebrows than his plans to create ride-sharing services that rely on self-driving Tesla cars loaned out by their owners during times when their cars aren't needed. It's not the kind of thing that consumers can expect to see anytime soon, but if and when they become available, such services could prove to be a big headache for ride-sharing leaders Uber and Lyft.

Musk claimed that when fully self-driving cars become a reality, users will be able to add their cars to a Tesla shared fleet by "tapping a button on the Tesla phone app and have it generate income for you while you're at work or on vacation, significantly offsetting and at times potentially exceeding the monthly loan or lease cost."

He added that such services effectively lower the cost of owning a Tesla to the point where "almost anyone" can afford one, and that Tesla will run its own car fleet in markets where the demand for such ride-shares exceeds supply.

As usual, some of Musk's claims warrant scrutiny. For example, he asserts the average car is only used for 5% to 10% of the day, and suggests it could be loaned out the rest of the time if it was autonomous. This estimate leaves out the many times during which a user will want a car to be instantly available, however, even if he or she isn't certain it will be driven. And Musk doesn't note the depreciation costs that come with the extra miles that will be racked up as a car is loaned out.

Nonetheless, when the technology is actually ready, it's not hard to see many self-driving car owners signing up for a service such as the one Musk envisions. And there's nothing about Musk's vision that suggests only Tesla would be able to provide such a service. Other automakers selling autonomous cars could create similar ride-sharing fleets, perhaps via joint ventures formed with rivals. Moreover, Tesla and other automakers could choose to run these services at cost, using them as an incentive to drive car sales.

The net result could be the commoditization of ride-sharing. Even after factoring in the spike in demand that might come as self-driving cars lower ride costs, enable cheaper delivery services and potentially make a greater percentage of the population willing to forego car ownership, the addition of hundreds of thousands, if not millions, of cars into the ride-sharing pool would greatly erode Uber and Lyft's main barrier to entry against would-be rivals. Namely, the massive network effect each company has created through its base of drivers and passengers.

Uber and Lyft are far from blind to the disruptive impact autonomous driving could have on ride-sharing. Uber has been testing a self-driving car (a modified Ford Fusion) in Pittsburgh, and has been frank about the long-term potential of self-driving cars to replace the need for paid drivers. Abd Lyft is partnering with General Motors (GM) - Get Report to develop self-driving taxis.

Moreover, even if Musk's vision becomes a reality, it's going to take a while. Google, which is perhaps further along than anyone when it comes to autonomous driving R&D and data-collection, is only aiming to bring a self-driving car to market -- likely with the help of one or more existing automakers -- by 2020. There are still plenty of regulatory and driver-liability issues to work out, and as Tesla itself has observed, there's a big difference between creating a car that can occasionally take over for a driver and one that can completely do so.

All the same, Tesla's plans should make those private investors who have given Uber a $62.5 billion valuation, a valuation that clearly assumes Uber will maintain a monopoly or duopoly-like position in most of its core markets, a little nervous.