Are Driverless Cars Fool‐​Proof? Not Quite

Randal O’Toole discussed the idea of safe, efficient, driverless cars in his book Gridlock: Why We’re Stuck in Traffic and What to Do about It and in this full‐​page Wall Street Journal essay in 2010. It wasn’t exactly a new idea — Norman Bel Geddes first imagined the idea 75 years ago at the New York World’s Fair of 1939 — but O’Toole was on the cutting edge of bringing it to more popular attention. And as he noted, one of the important benefits of driverless, or “self‐​driving,” cars is safety. As a driving‐​test site, citing British studies, says: “By far the biggest cause of road accidents is driver/​rider error or reaction, which causes 68% of all crashes.” The loss of control, the reliance on mysterious computers, scares many of us. But there’s good evidence that computers can guide both airplanes and automobiles more reliably than human operators.


But maybe not all human operators.


Meredith Shiner of Yahoo! News reports:

Scientists from Carnegie Mellon University on Tuesday brought a prototype of a driverless car to Washington in an attempt to show Congress that it could embrace a future devoid of man‐​made errors. 

And then Congress broke that car.

It was not immediately clear whether the mere proximity to the Capitol created the series of events that led to an emergency switch being flipped, causing the car to shut down, or if an actual member of Congress did it.…

In true Washington fashion, no one would take immediate responsibility for the developing car situation.

Okay, not entirely fool‐​proof. But getting there.


Update: NBC News reports: “D.C. Delegate Eleanor Holmes Norton hit the kill switch on the car before she was supposed to take a ride, and they couldn’t get it running again.”