On Dragons, Horses, Uncanny Valley, and Self-Driving Cars

You get something, you lose something. That’s how it always is with new tech. Digital photography lost quality and something more abstract—grain and physical paper-and-chemicals and emotion, perhaps—compared to film, and it took decades to catch up (and, of course, there will always be the purists who prefer film, ‘till kingdom come). Typewriters took away the uniqueness of handwriting, word processing took away the instantaneous feel of letters pressed into paper, eternally unchangeable. Records and radio took away something from live performances, and digital audio took away—crazily enough—the fuzziness that made the former feel “authentic”. I’m sure the wheel must have taken something away from brute-force work and, well, walking.

And thus it is with transportation. Horseless carriages were an obvious improvement over horses: they didn’t get tired, they were far faster, and they gave us smog instead of dung-filled streets (the former giving lung problems and global warming, the latter causing disease endemics. Ok, this point might be a toss-up.). But they were things. Horses are living beings that, almost, became one with the rider. You had an emotional connection with your horse—it was a real living and breathing creature that you’d feel bad for when it was sick. We mostly just feel annoyed when our cars end up needing fixed.

That very livingness gave something horse-based transportation something else that their technologically superior replacement never gave us: consciousness. It’s what gave us an emotional attachment to horses, but is also what kept people safe on them. Countless movie scenes have the rider dozing off or planning ahead or anything else while their horse trots on, no guidance needed. You’d start your horse off on the right path, then let it go on its own pace through the bulk of the journey. Get near your destination, and you’d switch back to more actively controlling your horse—or if there ever was danger, the horse would alert you to it and you’d take the reigns again, literally. And it’s not just horses: that very concept of the animal you’re riding being able to go on its own for the bulk of the boring stretches of your trip is always picked up whenever there’s a story scene with someone riding something intelligent: How to Train Your Dragon 2 illustrated this perfectly with Hiccup nearly falling asleep while Toothless continued flying to their destination, but then taking control again once there was danger. Self-driving vehicles, for the most boring stretches of the road.

This is what motor vehicles lost. Cars demand full attention, all the time, no matter how straight and boring the road. And yet, we’re terrible at paying attention all the time, gazing instead at billboards and the models of cars around us and—scarily enough—even at phones and newspapers and more. Our minds drift off, and then suddenly there’s something in front of us and we have to slam on the brakes. Surely countless accidents could be avoided daily if only our cars were as smart as horses.

As smart as horses. Imagine cars that could drive themselves on the straight stretches, when we feel save relinquishing control, and then let us take control again when we want. It’s not so hard to imagine, even: light-SciFi for years has shown cars that can drive themselves, but then let you take back control when you want to (of course, for action scenes when we’d want to root for the human who’s controlling said car). See the car in I, Robot, for example. It’s smart, but ultimately still a tool to be controlled by a human.

For there’s something human about being able to steer our own destiny, control our fate, and decide precisely how we want to turn into a curve and avoid the bad guy. We like tools that extend us—keyboards and musical instruments and cameras and hand tools. Automation is hard to trust—we all know that you can’t trust Google Maps absolutely, and are far better off supplementing the GPS with your own knowledge of the roads. How then can we expect people to trust fully automated cars, driverless cars that can take us to a GPS destination but don’t listen for our input otherwise?

That, I think, is why Google’s self-driving cars hit an uncanny valley for so many. We’d perhaps like a smarter car, a car that follows the lines and automatically brakes if we’re going to hit something. Perhaps we’d even accept “smart” cruse control that can drive the straight stretches, the boring parts of our commutes and road trips, for us, as the new Cruse is offering next year for $10k, and that we’ll undoubtedly see in many premium cars as a default feature within this decade. If there’s still a steering wheel, and if we’re still asked to make active decisions about our final destination—by hand, mind you, not just by abstractly punching in a GPS location—we’ll feel ok. The car will feel like a further extension of ourselves than the current gas pedal and steering wheel do. But if the car can only follow a GPS route, and must be fully automated—if it doesn’t give us any more control than entering an end destination—it’ll be far tougher to get us to accept it.

Tesla excites me, because it’s pushing the tech of cars forward in a mechanical way—it’s the most advanced dumb horseless carriage yet. Google Self-Driving Cars scare me in a way I know deep down they shouldn’t, because I happen to love driving, and love the control of being behind the wheel. There’s got to be a middle ground, where we can take back control at will, and give freewheeling directions to our cars—say, by saying out loud “Ok, Car, keep driving until we’re 1 KM from Exit 256, then let me know”. That’s a future of personal transportation I’d be ok with.

But I wouldn’t very much like being strapped to a horse that’s taking me to a destination, without any way to change the course or direct it. High-tech whiz-bang won’t change that.

Thoughts? @reply me on Twitter.