For the last few years people have been trying to sell me on the idea of autonomous race cars. I understand the appeal of Chris Andersons’s DIY Robocar series: they’re cheap, fit in your hand, and there’s trickle-up benefit to having different technologies race each other. What I don’t yet understand is the appeal of full-size autonomous cars costing six or seven figures racing each other, at least not based on what I’ve seen.
So I reached out to IndyCar driver JR Hildebrand to talk it out with Bryan and me on No Parking. JR’s a 10-time veteran of the Indy 500, which one might assume puts him squarely in the human racing camp, but beyond driving full-time for Dreyer and Reinbold Racing, he also moonlights at Stanford’s REVS Institute working on autonomous vehicle technology.
If anyone could make the case for autonomous racing, JR’s the guy.
My thesis was that people watch sports because they identify with the people. Communities identify with local teams, cities with their franchise, and countries with national teams. In sports like tennis or car racing, where a country might be represented by a single person, that human-to-human connection is even more important. We live vicariously through the people in the seat, pray for them when something goes wrong, and rejoice when they walk away. Remove the drivers, and there’s no one to root for and nothing at stake.
JR agreed, to a point: “There’s a human component to what we do that is super interesting. The car itself, the vehicle, whatever type and style it might be, whether it’s a 20-horsepower go-kart or it’s a 800-horsepower IndyCar, it’s an extension of you as a human. When everything is right, that’s what it is. There’s a great quote from [legendary driver] Graham Hill…I think the quote goes like, ‘I am the artist, the track is my canvas and the car is my brush.’”
He also argues that racing goes beyond “man in machine combat,” because advancements on the track expose larger changes in the automotive industry, and drive improvements of the cars we use every day. These advancements also drive development of the fully autonomous vehicles that will eventually be deployed on city streets.
OK, but none of that addresses the big question: would anyone want to watch two autonomous cars race each other?
“If you’re going to do autonomous racing,” JR replied, “I think that the conflict of human versus autonomous machine is the obvious, most interesting thing to see happen. But I think the machine being an extension of the human is still as far as the sport goes.”
I didn’t expect that, because it expounded on two controversial ideas: 1) that human race drivers will eventually see more driver assistance, which happened in the 1990s in Formula 1 until it fell out of favor, and 2) that a contest between human and machine is something he actually wants to see.
It seems obvious to me that eventually AI will advance to the point that no human driver can possibly win, and there wouldn’t be much entertainment value in it beyond a stunt. But of course we’ve not yet seen the racing equivalent of IBM’s Deep Blue beating chess grandmaster Gary Kasparov in front of an audience back in 1996. I’d be curious to see that… once. But just like Deep Blue didn’t kill human-versus-human chess as we know it, human racing isn’t going to evaporate even if an AI-driven car wins the Indy 500.
Why would it? People will always want to see people compete against each other, because that’s what inspires us to be better people. It’s human nature, which is something no machine can match.