We are all wired to respond to story, not statistics. News anchors rarely quote statistics – they tell stories. The most effective public speakers find the story in their material. Charities have long known that focusing on one child (or one puppy) raises far more money than the plight of thousands. Consider the following:
Joseph Stalin said, “a loss of one life is a tragedy, the loss of a million lives is a statistic”;
Incorporating “pain free” into Patients’ Bill of Rights made great anecdotal sense, but the opioid crisis it spawned wasn’t recognized until it reached epidemic proportions;
After 9/11, US skies may have been the safest ever, but thousands chose the “safer” option of driving, and so there were 2,300 more traffic deaths that year than would have been predicted.
The problem is that when it comes to dictators, drugs, or driving, “story-bias” can be deadly. I’ll leave dictators and drugs aside for now. I want to talk about driving (or more accurately, not driving).
Recently, a few colleagues and I had the great pleasure of visiting the Sensible Cities Lab at MIT (http://senseable.mit.edu/). This incredible team of scientists, researchers, engineers and other big thinkers are working on the future of urban life. Driverless cars (and boats) figure prominently in this future, and it got me thinking – the move to a truly driverless economy is taking way too long.
“Traffic accidents remain one of the top ten causes of death among young adults.”
Last year, more than 40,000 people died in traffic accidents in the US. Safety improvements have moved motor vehicle accidents out of the overall top 10, but it remains one of the top ten causes of death among younger adults. For males aged 17 to 23, it’s #1.
Which brings us to the very real, and admittedly very new, question of why people drive at all. The technical challenges of driverless cars are mostly solved, and while the social implications of this technology are still being sorted out, most of them are positive:
Taxi fleets in gridlocked urban areas like Manhattan could be reduced by 40%;
Parking requirements in densely populated spaces like Singapore could be reduced by 70%; and
We’ll need up to 90% fewer cars. Since the carbon footprint of making a car can exceed that of driving it, the positive environmental implications are huge.
“People talk endlessly about driverless deaths (all 5 of them) but far less about the 39,995 others that will happen this year”
The biggest positive impact is that motor vehicle deaths could be reduced by 90%. This would save the lives of a hundred people a day and eliminate the #1 cause of death in young men. The big barriers now are public policy, tort law, and “story-bias”.
When there’s a human at the wheel, there’s not much of a story, other than for those directly affected. When there’s nobody at the wheel, it makes for great story. That’s why people talk endlessly about driverless deaths (all 5 of them) but far less about the other 39,995 that will die this year.
False debates about malevolent, AI-driven cars “choosing” to save their occupants at the expense of pedestrians circulate. The fact is, in the very rare case where this situation arises, AI will behave exactly as the human that programmed them would – not by maximizing the survivability of one or the other, but by “satisficing” – picking the option that will result in the least harm overall.
The “who will we blame” debates roll on, mainly because it fosters an interesting narrative. Of the $99 billion a year that motor vehicle accidents cost us every year, most of it is paid by insurance companies in a no-fault way. When that number is cut by 90%, surely we can figure out the right answer.
Perhaps the largest challenge of all will be in the transition period where driven and driverless cars must share the road, and the main problem will be the humans, not the machines. My older son is a product trainer for a major car company. Speaking with him while he drove his nearly autonomous SUV down the highway, the biggest problem seemed to be the humans that kept changing lanes into the safe distance his LIDAR-controlled car was maintaining.
Driverless cars could reduce motor vehicle deaths by 90%
This weekend, we marked my younger son’s 21st birthday. We got dozens of text messages from family and friends wishing us well and expressing their love and support. The most deeply moving ones came from his friends, who still miss him as much as we do. According to the CDC, weekends in June are when teenaged boys are most likely to die in a car wreck. In Chase’s case, there were no drugs, no alcohol and (gratefully) no passengers involved. As his little sister might say, he was just “being a dumb boy”, driving teen-aged fast and taking teen-aged risks. He was 19.
Now that the technical issues are falling into place, it’s time for the politicians and lawyers to do their jobs. Let’s hope the workforce of the future has more of today’s 17-23 year olds, that 8-year old little sisters don’t have to deal with lost big brothers, and that friends spend Sunday afternoons in October watching football, not visiting a gravesite.
I’m betting this little sister (quite happily) will never attend driver’s ed, never have a driver’s license, and never even own a car. Most of all, I’m hoping she won’t know anyone from her high school class who died in a wreck just weeks before graduation – someone too many of us today know firsthand.