Wednesday, August 28, 2019


I've been staring at this self-driving car issue for years without having any clue how people could imagine it's going to work. It just isn't a problem that can be solved by making it incrementally better, especially when the fantasies are about "robotaxis," the hardest possible application. I suppose it's somewhat of a function of the roads that you are used to. Newly developed suburban road systems in areas of CA and AZ "seem" like solvable problems, but I can't picture these things traveling half a mile here in the urban hellhole, and not because the urban hellhole is hard driving for human drivers.
Mountain View-headquartered Waymo wants to expand the rideshare with its advanced self-driving taxi service. But if San Francisco test riders have anything to say about it, the previously Google-owned company still has a long way to go if its executives plan to compete with Lyft and Uber.
That they work as well as they do is genuinely neato. But neato doesn't mean useful.