Reading this article my conclusion is that "autonomous self-driving vehicles don't work at present". Ford management had the belief that they do work and therefore wanted to press ahead with a go-to-market plan. Google management knows they don't work and so back-pedaled. Deal fell apart as a result.
That conclusion is almost certainly incorrect. Internal communications at Google indicate strong confidence in the approach and viability of the cars. There was constant crowing about hundreds of thousands of miles driven autonomously (with the only at fault accident being the car committing the egregious sin of expecting an overtaking public bus to yield for an indicated merge). A recurring internal criticism is that they have been ready to go to market for a while, but are proceeding with extraordinarily caution. Google probably wants a repeat of the Personal Computer market: open, agile, competative. There's a hint of this in the way they structure deals (quietly helpful, non-exclusive). "You want the technology? Sure. You want exclusivity? Nope." If Uber had asked for a technical partnership instead of straight up stealing the tech they'd probably have gotten what they stole and more with a smile and a high five.
