Uber fatality sends self-driving technology back to the drawing board where it belongs
On March 18, a woman was killed in Tempe, Ariz. by an Uber self-driving car after she stepped outside a crosswalk in front of the vehicle. The vehicle was in automatic-driver mode, and there was also a “pilot” inside who could have stopped the tragedy. But police said he had reached down to pick something up off the floor and wasn’t paying attention.
Two humans, the victim and pilot, obviously share blame, but the principle culprit was the robot car which was ill-equipped to avoid a collision with anything. In this case, the collision was with a human being and ended her life.
But this fatality, and another involving Tesla automatic mode, is hugely significant to the auto and tech worlds who have been racing to get autonomous vehicles approved for mass adoption. But that notion is now delayed for years as the accident, among many non-fatal mishaps, has sent computer scientists and law enforcement officials back to the drawing board.
Ironically, Silicon Valley proponents boasts this is the technology that can eliminate the 30,000 car fatalities annually. But only if it is foolproof, which is not the case yet.
Such a delay will give society a pause to examine the fact that autonomous vehicles and drones will create mass technological unemployment. Being a truck driver happens to be the most common job in 29 out of 50 U.S. states, and represents an estimated one-in-eight in Canada. An estimated 3.5 million drive trucks across these countries long distances and millions more drive other vehicles from taxis to Uber cars, vans, delivery vehicles, and buses.
For this army of people about to be automated out of their livelihoods, any delay is good news. It’s also vindication for the powerful Teamsters union and others who have been lobbying against allowing these vehicles on public roads even for testing, much less for mass adoption.
“I’ve driven 4 million miles and yet every day I head out there’s always some new situation I have to deal with,” a truck driver told the media last week. “Can it [computers] really distinguish between a deer and a child and always make the right call?”
Drones represent another blow to employment when they begin to deliver people and packages short and long distances. Fortunately, regulators such the Federal Aviation Administration require all drones to get permission to fly as they devise a ubiquitous air traffic control system.
The Teamsters opposition to autonomous trucks represents the first shot fired in a looming political war against mass automation that will monopolize politics this decade and next. The reality is starting to sink in that automation products are and will ravage the middle class, notably robot-cars as well as Amazon’s voracious platform which is eating the retail sector alive.
The second most common job in America is retail clerk or cashier.
Amazon, in just 24 years, has gone from a website offering books to revenues that now surpass all of America’s department, grocery store and restaurant chains combined. It’s technological efficiency and growth does not merely raise questions about unemployment but also about excessive concentration of economic power.
Clearly, cars that drive themselves are on hold and should be. These vehicles can eventually be a godsend to the elderly or disabled or to public transportation. But leaders must do a cost benefit studies before they allow technologies to lay waste to workers.
Besides, machines are only as intelligent as the least intelligent human being that designed and manufactured them. Will man-made machines be better? And will the machines support the drivers, pilots, and clerks they replace?