Whither Self-Driving Cars?

Nobody asked but …

Now a self-driving car has killed a human being crossing the road.  Some of the news following on the occurrence has imputed negligence (human error) as part of the chain of events.  I’m thinking the story is similar to that of  Three-Mile Island.  In retrospect, the incident at Three-Mile Island, near Harrisburg, PA, put an end to nuclear power plant proliferation in the USA.  Although it was, in hindsight, a minor event as far as technology goes, it introduced doubt which dispelled the popular enthusiasm for nuclear energy in the power business.

In neither case was the technology directly to blame.  At Three-Mile Island, a human misinterpreted a user message on a console, while in Phoenix, somewhere in multiple decision chains, humans had made seemingly innocuous moves that had accumulated in a lethal event.

Looking back, I never thought self-driving cars were a good idea.  The probabilities, mathematically, are overwhelming.  Although hindsight is 20-20, I must say that my skepticism was in foresight.  But I am not trying to squelch innovative thinking.  I believe that the real problem was overenthusiasm.  First blame falls on the shoulders of a pop science media, the members of which are deluded by a hefty pop response — people have a furious love-hate relationship with labor-saving devices.  Second blame goes to a gullible public, who confuse popular response with likelihood.

I don’t want to be too quick to say last rites over self-driving cars, but fallibility is the Achilles’ heel of the illusion of infallibility.  And ideas are far more dependent on popular acceptance than they are upon usefulness.

— Kilgore Forelle