Keep calm and let the car drive itself

[CNet] Several times in the past year, I’ve climbed aboard a jet airliner, and it has carried me to some far-flung destination abundant in sunshine and balmy temperatures.

Sunshine and balmy temperatures are two things we don’t take for granted here in the UK, but another aspect of the sentence above, we do.

“It has carried me,” for example. “It” being the aircraft, of course. And using the singular neuter pronoun “it” refers to something we’re all acutely aware of these days, which is that it’s quite possible for an aircraft to conduct most of its journey without input from the pilot and co-pilot.

By and large, our journeys aboard modern aircraft are left largely in the hands of a computer. And we just sort of…trust it.

In fact, we trust computers with a great deal these days. We trust them enough to store our priceless family photos on, and to buy expensive products across the Internet, with little more than encrypted bits of data hiding our valuable bank details from prying eyes.

We also trust that computers will coordinate systems that let those expensive products get to the place they’re being stored in the first place. Or indeed, that the products will be built in the first place, using computer-controlled machines.

You’ll already have guessed where this is going: Why, given all that trust put in computers, is it so hard to trust cars driven entirely by computer?

The technology is certainly developing quickly. Quickly enough that Florida, California, and Nevada all allow driverless cars on their roads.

Just think about that — it is legally possible for a car to make a journey with no human input whatsoever, on the public roads of three US states. The laws were partly granted so that autonomous car advocate Google could test its own vehicle on the streets.

Google’s self-driving Toyota Prius has now driven more than 300,000 miles itself, accident free.

Well, not entirely accident free — one of the car’s technicians, while driving it manually, managed to cause a five-vehicle crash. And on another occasion, the driver of an entirely manually controlled vehicle managed to run into the back of the driverless car.

The pattern here is that in nearly a third of a million miles, the computer hasn’t made a single mistake — yet humans have.

Humans make mistakes all the time. We worry about computers driving us around, yet we’re happy to trust fallible, distracted, emotional collections of flesh and bone to do the same.

Leave a comment

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.