scorecardresearch Skip to main content

Self-driving cars: Overcoming the fear of something going wrong

People who heard about my deep-dive research into driverless cars tended to react in one of two ways: either when-can-I-get-mine eagerness or can-we-really-trust-this wariness. I know where both camps are coming from.

When I sat in Google headquarters with Chris Urmson, the 38-year-old head of the company’s self-driving car team, his determination to tackle the remaining hurdles in self-driving technology was so obvious that the future immediately felt closer. He spoke evangelically of the explosion in productivity that our society will see if commuters are freed from the massive time suck of sitting behind the wheel on congested roadways.


When I asked him how old kids would need to be for parents to feel comfortable putting them alone in a self-driving car, he replied, “How old are kids when parents allow them to take the subway in New York City?”

I had to admit that being freed up from having to shuttle our kids around to school and endless activities sure sounded liberating. This route could also have the benefit of injecting more self-sufficiency into the lives of suburban kids, giving them the kind of important life skills that many city kids get from having to find their way around public transportation systems at a young age.

And this self-driving enterprise appears to be in very good hands at resource-rich and talent-heavy Google. (Upon arriving at its headquarters in Mountain View, Calif., I was struck by just how much more massive Google has gotten since my last visit a dozen years ago.)

To their credit, Urmson and his Google self-driving team had scrapped an initiative to offer automated highway driving, which by now would have been up and running. They recognized that any plan requiring human drivers to serve as the backstop for an automated system wouldn’t ultimately improve safety. So they switched gears to go for the “moonshot” of eliminating the need for human drivers altogether.

Despite how tantalizingly close this self-driving future appears to be, I also came to see the legitimacy in some of the worries that critics have. This has nothing to do with the recent admission from Google that its self-driving cars have been in a dozen accidents over the last six years.


Even though critics gleefully pounced on that news, the reality is that’s a relatively small number of mostly minor accidents given that the Google fleet has covered more than 1.7 million miles, and none of the accidents appear to have been caused by a self-driving vehicle. The biggest problem for driverless cars is that they will still need to share the roads with human drivers – people who do insane things while behind the wheel, like use both hands to eat a bowl of cereal or put on eyeliner or play the trumpet.

There are important ways in which self-driving technology will have to improve before we can feel comfortable that it’s ready for widespread use. Probably harder to overcome is the deep-seated fear some people have of being left stranded by this technology, even after it seems ready for prime time.

Think of an instance where an ATM went haywire and not only failed to dispense your cash but also swallowed your card. That’s a really annoying experience. But the stakes would be dramatically higher if a snafu like that happened to your self-driving car, especially if you had just entrusted it to drop off your 10-year-old at soccer practice.

I gained a greater appreciation for this fear while I was reporting my driverless car story in both Singapore and Silicon Valley, thanks to a not-ready-for-prime-time performance by a current darling of the tech world, Uber.


That mobile-app, on-demand transportation service, which has a ridiculous $50 billion valuation has by now so fully penetrated popular culture that it’s marbled itself into the repertoire of lazy late-night comedy writers. I decided I should finally join the crowd.

But I soon learned that Uber no longer supported the older operating system on my iPhone. Uber does, however, offer assurances that people with older phones can still use its service through its mobile site. For me at least, those assurances turned out to be empty.

In my frustration, I hit the “customer support” button, but got a message saying that the Web page was down. I found a direct e-mail address for customer support and detailed my frustrating experience, adding, “No idea if this e-mail will even get to a human. But if it does let me know.”

Three hours later, I received a cheerful e-mail from a woman named Shilonda in customer service, who wrote, “I’m a human and happy to help get you set up with Uber. :)” Unfortunately, she simply sent along a link to the app that I already knew didn’t work for my operating system.

After I explained this in a reply, she offered me a link to the mobile site.

“I tried that route before and just tried it again,” I noted in my reply. “But both times I got an error message telling me to download the app, which, of course, I can’t do.”


I didn’t hear another word from anyone at Uber until nearly a full day later, when Shilonda wrote, “Sorry for the late response. I was off work when this message came in.” She asked me to send her a screen shot of the error message I was seeing.

By then, I had already flown from Singapore to San Francisco, taking conventional taxis in both places. I was tempted to write back to Shilonda, explaining that I assumed I had been communicating with the support team of a $50 billion tech company, not a single employee whose days off I need to concern myself with.

Instead, I just let the correspondence die, thankful that I hadn’t put myself in a position where this celebrated tech giant’s missteps would have left me stranded on the side of a road.

Neil Swidey is staff writer for The Boston Globe Magazine. E-mail him at and follow him on Twitter @neilswidey.