The reality of autonomous vehicles is already here.
Good morning. It’s Cliff Leaf here, stepping in for Alan.
The moonshot is over. That was a key takeaway from Fortune’s fascinating conversation last night with Waymo CEO John Krafcik. Sitting down with Executive Editor Adam Lashinsky, at our annual Fortune Brainstorm Tech in Las Vegas dinner, Krafcik said that the real man-on-the-moon moment in driverless cars occurred in 2015, when a truly autonomous vehicle—one with no driver controls or emergency stop buttons—transported a human passenger through the streets of Austin, Texas. That feat took Waymo—now an independently run company within Alphabet—six years to accomplish.
Four years later, roughly 100 Waymo-converted Chrysler Pacifica minivans now driverlessly drive passengers through the streets of Phoenix, Arizona, each day. Some 1,500 active monthly users hail the vehicles via their Waymo apps and take them wherever they’d like within a service area as large as San Francisco. (The prices for the rides are comparable to those for Uber and Lyft.)
The sheer everydayness of this Jetsons–like experience—even if it’s limited for now to parts of Phoenix—was striking to hear. But as Krafcik told the Fortune audience, it took miles to get to this point; more than 20 million miles, in fact, of real-world road testing, with half of that coming in just the last year or so.
Backing up those 20 million miles of rubber-on-road have been tens of billions of computer-simulated miles, which have helped Waymo improve its algorithms, Krafcik said. As for the investment thus far—which the Waymo CEO pegged in the billions of dollars—“it would be hard for any company other than a company like Google to deliver,” he said.
So how, Adam posed, will the company make back those billions? Four ways, Krafcik said. There’s ride-hailing, then urban delivery. (“Using those same vehicles that we’re employing in Phoenix to move goods from point A to point B.”)
Then long-distance autonomous trucking. Waymo is generating revenue now through autonomous ride-hailing, but Krafcik expects commercial trucking to follow. Trucking “is a bigger market opportunity,” he said. And finally, Waymo can sell tech to automakers that want to develop their own applications.
The problem with thinking of such ambitious quests as “moonshots,” said Krafcik, is that it encourages us to frame the challenges in the wrong way. Missions like the original Apollo effort, he said, had a high probability of “some catastrophic failure”—which isn’t exactly an appropriate risk profile when you hope to deliver millions of something every day (as in driverless rides). “So we had to move beyond moonshot thinking, which got us to 2015, to this new sort of thinking—where we’re scaling, where we’re able to do this over and over again with very high confidence of perfect outcomes every time.”
Sure, it will require billions of dollars of additional investment. But then, what else are you going to spend all that Alphabet cash on?
More news below.
Clifton Leaf
@CliftonLeaf
clifton.leaf@fortune.com
