I will be neither surprised if they do nor if they don’t. Tesla have ~ten billion miles of experience, so at this point I think the question is “how long will it take to figure out a data-efficient algorithm?” rather than “can it be done?”

Back in 2009 I was expecting this to be available before 2019, and I didn’t have in mind anything as limited as either the “motorways only” autopilot of Tesla nor the (single?) heavily mapped city system of Waymo.