Baby You Can Drive My Car

With apologies to The Beatles, of course.

There’s a lot of discussion about autonomous cars in the news over the last year or so, and there’s plenty more on the horizon.

There are many, many roadblocks (so to speak) before self-driving cars become a reality. The sad part is that the technology won’t be biggest.

I’m convinced that the technology supporting self-driving vehicles will be ready long before consumers are ready to accept it. I think it’ll boil down to two significant, and completely psychological, issues: understanding and control.

We fear what we don’t understand

The ability for a car to drive itself is beyond most people’s comprehension. Seriously, it might as well be so much magic. We have no mental model that allows us to even comprehend something anywhere near the complexity that it represents.

What we understand is that driving is hard and we each have to take significant time to learn how to do it safely. Even then people still get it wrong to degrees varying from annoyance to death.

Yes, driving kills people. That’s scary, but we accept it; mostly because we understand what it takes for us to give it our best shot. There’s no way to really “get” exactly how a car would do that for us.

Besides, we rationalize, it’s the other people on the road that are the real problem, right? The ones we have no control over?

We fear what we cannot control

It’s bad enough that we can’t control “the other guy” on the road. At least, however, we have some semblance of control over our own situation: we’re the ones driving our car. We’re the ones in control. We have faith, justified or not, in our own abilities to both drive safely as well as avoid disaster when situations arise.

Self-driving cars ask us to give up control. We’re asked to give control to, not another person, but a device. A device that, as we saw above, we can’t understand. A device that, at least initially, we have no reason to believe will be any better than we are at driving safely and avoiding disaster.

In fact, with no evidence to the contrary, the closest parallel we’ll look at will be how often our personal technology – computers, mobile phones, and the like – give us fits, which is often. It certainly doesn’t help that we talk of computers “crashing” – even the terminology involved breeds fear.

The result is that we’ll feel that giving up control will put us at more risk than doing the driving ourselves.

And that’s scarey.

Fear leads to poor decisions

Various safety systems in advanced vehicles are already making the roads safer for those who employ them. There’s actually a very strong argument that fully autonomous driving will be safer – probably significantly safer – than humans doing the driving.

Your car doesn’t get tired. Your car doesn’t fall asleep. Your car doesn’t drive “impaired”. Your car is always paying attention – even while you change the radio station. Your car’s ability to drive will not decline with age. Add self-driving cars that communicate and coordinate with each other in real time at high speed, and the potential for safety improvement is dramatic.

I’m certainly not saying that technology is without risk. People will die in self-driving cars as well.

The problem is actually a very simple one: which approach will be safer, and when? Which approach will kill and injure the fewest number of people?

Technology will shift the answer to automation long before people are willing to accept it.

What that implies is, itself, scary. It means that people will actively continue to choose a path that will result in more death and injury than is necessary.

All because of unwarranted, albeit perhaps understood, fear.

Fear of progress

I know, I know, we’ll get past it. We always do.

This is really nothing new. I’m sure that every major new technology has faced similar fears. Ignorance (meaning a lack of knowledge or understanding) has lead to many periods of poor decisions in the past, and I’m certain that this is no different.

As for me? I’m ready.

My new car has the hardware to be fully autonomous. Sometime within the next year or two the software will be ready.

People, and legislation, on the other hand, will take longer. Perhaps much longer.

That wait will be frustrating.