This interview with Toyota autonomous vehicle developers at IEEE Spectrum supports what I’ve been saying for years about autonomous vehicles rather neatly. Short version: don’t expect a fully autonomous vehicle to pull up on the road next to you any time soon.
The foundations for what the interviewees refer to in the article as “deep learning” were there sixty years ago. There have been no basic developments since. What there has been is hardware that’s small and cheap enough to do things that were impractical sixty years ago. As I sometimes put it my smartphone has more compute power than all of the computers I’d ever worked on until 1980 put together.
To get around the problems you described, watch for a push to try and curtail or even ban human-steered vehicles for safety’s sake. After all, people behind the wheel kill tens of thousands of others just in this country every years, worse than what happens in some pandemic years. And it would eliminate the drunk driving problem. To program the fully autonomous fleet though you’d have to eliminate other major obstacles such as bicycles, people walking, etc. etc. But those are just mere details on the road to everyone’s goal of a totally safe and stress-free living environment.
Cynical in the Morning
Air taxis will beat them to it.
Here’s just one:
https://www.theverge.com/2020/6/8/21284140/lilium-flying-car-evtol-funding-investor-baillie-gifford