AILSA CHANG, HOST:
Dr. Jonathan Slotkin had witnessed this scene too many times at the trauma center where he works.
JONATHAN SLOTKIN: They're wheeling in a teenager, and this one's been ejected through a windshield and found 30 feet away, face down in the dirt, and we all recognize there's nothing we can do. And so I'm thinking to myself, how do we let the equivalent of one plane full of people, more than 100 lives, be lost every single day as the cost of driving? If this was a disease, we would have declared war.
CHANG: And right around this time, Slotkin had been looking at data released by the driverless car company Waymo. This was its so-called 100-million-mile data set.
SLOTKIN: And I looked at this data and said, God, this almost looks too good to be true. If we pursue this right, in my opinion, we face the possibility of actually eliminating car accidents as a leading cause of death in the United States.
CHANG: Slotkin described this epiphany in an op-ed for The New York Times, writing, quote, "while many see this as a tech story, I view it as a public health breakthrough." Dr. Jonathan Slotkin, welcome.
SLOTKIN: Thank you for having me, Ailsa.
CHANG: So you wrote that this Waymo traffic accident data led you to conclude that autonomous vehicles could be a public health breakthrough. What did the data show exactly that convinces you of this?
SLOTKIN: What I saw was that we had greater than a 90% reduction in the most serious types of crashes that we see. So these are pedestrians struck, T-bones in intersections, which are some of the worst injuries we see in the trauma bay. So it really looks like, Ailsa, that if we pursue this right now going forward, we could eliminate car accidents as a leading cause of death in the United States.
CHANG: OK. But if we are talking about safety, there have been recent reports of Waymos driving pretty aggressively, like making illegal U-turns, neglecting turn signals, illegally passing stopped school buses. And just last week, personally, a Waymo cut me off in moving traffic right here in LA. And you even mentioned in your op-ed a few fatalities. So shouldn't any of that be a cause for concern?
SLOTKIN: Well, sure. Let me point out for those that didn't read it, those fatalities were both not the fault of the Waymo, OK? And there was one serious injury as well that was also not the fault of the Waymo driver. But certainly, these are questions we have to ask, and they're all fair questions. The school bus thing, that's serious. We can't have these things passing school buses. And what I understand is they identified a gap where the cars were getting confused, and they issued a recall to update...
CHANG: Exactly.
SLOTKIN: ...The software. But I think let's look at the alternative. The alternative is human drivers illegally passing school buses hundreds of thousands of times a year because they're impatient. And here we have a robot that makes a mistake, and it can be rapidly remedied with a software update. So I think we accept humans being reckless as the price of mobility, but we want the robots to be perfect. And somewhere in the middle there, I think, Ailsa, is where we should land.
CHANG: I think, for me, any sort of lingering discomfort I have with autonomous vehicles is just this inherent trust I have in human supervision. Like, I want a human to be able to intervene in potentially risky situations. Is that illogical?
SLOTKIN: It's a natural human tendency, I think, to be afraid of the things we perceive we can't control, and we run into that in health care a lot. This I-can-drive-after-two-drinks thing some people have - well, that's the perception of control, but I'm afraid of the robot that actually is statistically better than me, but because I can't control it. And I think, for me, that's where the data needs to start to take us and say, well, look, this is actually better.
CHANG: Well, let me ask you this. Self-driving cars are still a pretty slim minority of cars on the road, even in cities like LA, where I am, where these cars seem to be everywhere. So short of full adoption, is there truly a public health benefit to having even just a few self-driving cars on the road?
SLOTKIN: Yeah, so this gets to an interesting question, and no one actually knows the complete answer, but there is some published data that shows that as these scale, certainly early in the adoption, you may get disproportionate benefits even relative to how much has been adopted. So let's make up a number and say, we could have 30% adoption but actually decrease fatalities by 40%. And that gets to things like network effects.
And, Ailsa, we haven't even scratched the surface really on hive behavior between these vehicles where they begin talking to each other in an increased manner. They relay things about safety challenges ahead and they behave better around each other. So there is evidence that even early in the adoption curve, we will see safety benefits.
CHANG: Dr. Jonathan Slotkin, thank you very much, and happy riding to you.
SLOTKIN: Thank you, Ailsa. Thanks for having me.
CHANG: And we reached out to Waymo about the road safety violations mentioned in this story. The company responded in a written statement that said, in part, safety is fundamental to everything that we do. The data shows we are improving road safety in the communities in which we operate.
(SOUNDBITE OF MUSIC) Transcript provided by NPR, Copyright NPR.
NPR transcripts are created on a rush deadline by an NPR contractor. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.
