Many factors in the self-driving industry have happened as I anticipated. Consortia have formed, solidified, and dissolved. Lawsuits have popped up. A great bull market for technicians and managers with top knowledge has materialized. And technical progress, along with investment and business effort, has marched forward. So, when looking at updating our employment and technology-saturation projections, what less-expected events from the past year need to be incorporated?
First, the crashes, especially the fatal one, received more negative reaction than I would have thought. They were not that disturbing, given the extreme situation with the pedestrian in the dark and most others owing little or nothing to the vehicle’s driverlessness. Yet it was worthwhile for companies to see how people would react, as they were certain to occur sometime.
Second, Uber, with its business recklessness, has muddied the autonomous waters. The Arizona crash was just one example of why it does not seem to have enough due diligence for this field, and it is too easy for people to conflate its mistakes with the work of far more reliable firms.
Third, it is possible, though it is hard for me to assess from outside the field, that technical progress has been a shade slower than estimated.
Fourth, already have emerged some super-strong players, particularly Waymo and Aptiv. They could still end up Stanley Steamers to someone else’s Toyotas, but the chances are edging down.
Fifth, remote human control, which two years ago I made a stage of my own set of automation levels, is now getting press mention. In the National Highway Traffic Safety Administration’s Society of Automotive Engineers scheme, which I continue to use as below, it would fit in well as an option for level 4, or high but not full automation, or near the end of level 3, which still requires a driver to be available.
Sixth, publicity, toward understandings of how many kinds of autonomous vehicles and their interiors there will be and on how wide a range of life-changing possibilities they will end up having, has started. Fall’s New York Times magazine section was especially effective.
Seventh, it now looks likely that there will be real regional differences in driverless-vehicle acceptance within the United States. That will be a phase lasting as long as ten years and will for that time cut into overall proliferation acceptances.
Eighth, some other countries, particularly Russia and China, have done their own autonomous research and development, but their closed communication styles and lack of vetted progress make it difficult to consider their efforts world-class.
With all these things considered, here are our new projections:
For definitions of the levels, see the original NHSTA document at.
Stay with the Work’s New Age blog for at least quarterly updates on the progress of driverless vehicles and its effect on jobs. We will publish our fourth forecast next summer.