I don't believe that this is actually happening for at least 50+ years. And by that point, indeed, I won't be driving myself, but for slightly different reasons.
I work in the same building as Waymo. They’re very serious about self driving. I don’t know anything about their operations internally or how well they can handle inclement weather, but I think 50+ years is overly pessimistic. Self driving cars work today in clear driving conditions. Personally I feel like it may be more than ten years until we see them everywhere, but maybe it would be less. 50? That’s just silly. I don’t think it’s such a technical challenge we couldn’t solve it sooner.
I mean in the 60s there was a famous project where the idea was to "solve" image recognition, with an estimate of few months tops. It's 2018 and the best of the best image recognition algorithms will say with 99% confidence that a sofa in a zebra print is in fact, a zebra, and our technological leap from 60s is just unimaginable. We have so much more computing power, yet we can't crack it fully. I feel like autonomous driving is the same - it's going to be amazing in some general cases, but it's going to fall apart at any normal situation that can happen on the road - and it's those edge cases which will take decades to work out fully and reliably. So yes, I genuinely don't believe that we will see a fully autonomous vehicle that can work in any conditions, on streets of San Francisco as well as rural Ukraine equally well and without any human input.
Sure, they will not be deployed universally but they will soon be common for interstate trucking and taxis. Also expect to see different adoption in the US than Europe and Japan (like Norwegian and Japanese cell networks vs the US). Needless to say, I agree rural Ukraine may have different needs than San Francisco. But that has nothing to do with adoption in San Francisco.
Adoption does not have to be 100% to be significant.