Almost 3% of all American workers drive for a living. These include truck drivers and others involved in transporting goods and people. The rise of e-commerce has resulted in an increase in the number of driver-driven jobs, particularly as Amazon has hired its own fleet of package deliverers through its Flex program to augment its capacity beyond what UPS, FedEx, and the US Postal Service can handle. And, of course, Uber and Lyft do for people what Amazon’s various systems do for packages. There are more and more drivers moving more and more people and stuff more and more places.
Consider what will happen to these jobs when the long-promised vision of reliable and safe autonomous transport becomes a reality. Trucking is the largest employer in some states, reaching 5% of the working population. Nationwide, 3.5 million people work as truckers, and nearly 8 million people work in the trucking industry in some capacity. That does not include people who use their own vehicles to transport goods and people. Uber alone employs over 2 million drivers. What happens when the marriage of computing and automotive technologies gets to be so good that people are no longer needed to provide these services?
Are the computer scientists and engineers working on these technologies thinking about that question? One might argue that they need not be. After all, other technological revolutions eliminated jobs that used to be commonplace. You won’t find too many job openings for coachmen today, after all, so you might conclude that today’s horseless carriage conductors are getting their comeuppance. Every technological age has ushered the demise of some segment of the job market.
But this revolution seems different. Intelligent automation was once seen as the key to the four- or even three-day workweek. But one has to wonder if, given the unprecedented access to virtually unlimited quantities of data, the kinds of quantities that can fuel truly powerful artificial intelligence, the degree to which automation has advanced today might actually eliminate the workweek altogether for vast portions of the world’s population. While there is certainly no consensus among economists and policy experts, some of the forecasts of the future job market predict profound proletariat carnage. Indeed, one study predicts the loss of up to 800 million jobs.
This week, Amazon unveiled Scout, its “adorable delivery robot“. Amazon has long focused on making the so-called last mile of its deliveries more efficient, even suggesting back in 2013 that packages would soon be delivered by drone. Stalled by increased regulation of unmanned aerial vehicles, Amazon is now going full steam on more land-bound transport. About the size of a small cooler and certainly looking way cooler than that guy delivering Alexas and toilet paper out of his Kia, Scout is smart enough to deliver your goods right to your door, without running over your dog.
Unfortunately, while it may spare Rex, it might just run you right out of a job.
Repeat after me: just because we can build something … doesn’t mean we should. Sometimes the technical challenges are easier to solve and less consequential than the social ones. As we continue to push the envelope of what is possible, particularly in regard to data-driven artificial intelligence and autonomous control, we have to mind the consequences. Not everything that can be made possible should, no matter how “adorable” it might look.