Image of data-gathering by self-driving car on highway. (KitGuru)

The Atlantic has a very interesting article, “How Self-Driving Cars Will Threaten Privacy,” by Adrienne LaFrance. It actually ran a couple of years ago, and looks forward to the convenience of life with a self-driving car. The car will listen to your conversations and take your unwitting advice by, say, putting on your grocery list a type of beer you praised on the phone to a friend, or, without having to be told, adding a coffee stop on your way to work. Nice!

But there’s a darker side to all this, too. Let’s rewind and take a closer look at your commute for a minute.

There we were. The car picked us up. We wanted coffee. It suggested Peet’s. But if we’d stopped to look at the map on the screen when this happened, we might have noticed that Peet’s wasn’t actually the most efficient place to stop, nor was it on your list of preferred coffee shops, which the car’s machine-learning algorithm developed over time. Peet’s was, instead, a sponsored destination—not unlike a sponsored search result on Google. The car went ever-so-slightly out of the way to take you there.

The analysis by LaFrance goes into chilling detail about how the desire for profit might cause your car to edit your life.

Still, a lot of people will decide to put up with self-driving surveillance of this sort. We already put up with it online, allowing Google or other busybody behemoths to place ads for things they know we want in the articles we read on our laptops. And who knows what else. Better living through algorithms!

However many billions or trillions that Google, GM, and other corporations (including Uncle Sam) plow into self-driving cars, I don’t think they are in the cards. Our brains function faster and better behind the wheel using intuition developed over many thousands of years than any computer. Maybe computers can do a lot more calculations at once, and maybe they don’t drink and drive, but they can also be relied on to break down in little ways that don’t matter so much on our desktops but might matter a lot more hurtling down the highway at 65 miles per hour. The amazing thing is not how many accidents we have today but how few. Thank your brain.

So I think that after a while, after more serious testing ramps up and demonstrates the beauty of our natural onboard computers, we will shrug our shoulders and admit … no can do. Hopefully, we’ll have some useful things that are spun off by the research, like Tang was spun off by the space program.

But if it does happen – and this is one of many good reasons for it not to happen – what I worry about most is if, one day when self-driving cars are installed, I write a piece describing how we still have not been able to work out the 3D Rubik’s Cube of changes that will be required to reconcile the massive dislocations in technological, commercial, infrastructural, social, civic, economic and other systems that emerge over time because of the self-driving car. Suppose someone at Google decides to get the Ministry of Truth to issue a warrant for my arrest. All they have to do is plug a new itinerary into my car’s computer and instead of driving me to Peet’s, my car (or some corporate fleet’s car) will drive me straight to jail.

Nah. This cannot be allowed to happen.