We may earn a commission from links on this page.

Despite decades of interest in self-driving cars, the first large-scale attempt at producing them yielded little more than a series of broken promises. Now, autonomous-vehicle companies are giving the concept a reboot, learning from past mistakes and charting a new and, they hope, more successful road ahead.

Quartz technology reporter Michael Coren, with help from Hong Kong reporter Echo Huang, explored the past and future of AVs in a field guide published this week. On Aug. 22, Coren joined tech editor Mike Murphy—Mike and Michael take the mic, we planned it that way—for a members-only conference call to discuss the state of the industry. You can watch the full video above or, for the major takeaways, keep reading.

So… what’s the issue?

The idea of autonomous vehicles is cool, yes, but there’s a larger question: What do self-driving cars actually solve? They’re convenient, sure, but the breakthroughs in artificial intelligence that enabled much AV development didn’t necessarily prioritize safety. At no time was that more apparent than when one of Uber’s self-driving cars killed a pedestrian last year.

Give me the key concepts.

It’s a design problem, not just a data problem. Our existing infrastructure and behavior are the training data for AVs, but if machines are learning from our current system, we’re effectively doubling down on today’s safety and pollution problems. That doesn’t mean autonomous vehicles are a non-starter, Coren says; we just need to find better ways to integrate them into larger transport solutions. Right now, the most immediate proposal is to integrate AVs into freight transportation. One human could lead a platoon of one to three self-driving trucks, which could change America’s supply chain.

A tale of two cities. The marriage of Detroit and Silicon Valley will change the future of autonomous vehicles. Silicon Valley has a culture of moving fast and breaking things; Detroit has a culture of knowing it builds machines that can be deadly. While Silicon Valley may hold a competitive advantage, it doesn’t historically have the foundation of safety and thoughtful engineering necessary to build successful AVs. The companies that can figure out how to unite those cultures will be best positioned moving forward.

Regulation rests on philosophy. When an autonomous car crashes, who is at fault? The car? The programmer? A Quartz Member was curious about how to deal with the ethics of AVs, but the answer isn’t clear. In fact, as Coren noted, a panel of philosophers has discussed this very issue—arguing a car following Buddhist precepts might make different decisions than one abiding by traditional Western or Confucian concepts. Coren suggested that in the future, self-driving cars designed for China, Europe, and North America might abide by somewhat different ethical standards, though it’s also possible the industry will end up looking more like the airline industry than today’s car industry, with a baseline standard of safety. For now, it is “a free-market melee,” with no expected regulation from the Trump administration based on industry insiders’ conversations with the White House.

Give me the best quote.

“There is Elon Musk, there is everyone else.”
—Michael Coren on approaches to safety and competition in the autonomous vehicle industry

So what are next steps?

Keep an eye out for autonomous vehicles in the near future. Simply dropping self-driving cars into cities would be a terrible idea—implementation will require a whole new set of policies and city planning. But realistically, self-driving vehicles can be used for small-scale transit between certain places. For the 2020 summer Olympics, Japan is expecting to have the biggest display yet of autonomous vehicles. And freight systems will see autonomous vehicles in use sooner rather than later, Coren says.

Invest in transportation that works. Though self-driving cars won’t solve all our pollution problems and transportation anxiety, cities can invest in buses, trains, scooters, and bikes because those are things that work. What we should be investing in is smart, high-density transportation systems.

How can I learn more?

Read these recent studies and books (curated by Coren) to get up to speed on the latest in autonomous-vehicle technology.

Watch what it’s like to ride in a Google self-driving car (or a WeRide AV in China).

Listen to this story about the ethics of self-driving cars and what people decide when determining how a self-driving machine should act.

When’s the next conference call?

I’ll do you one better: The entire conference call calendar can be accessed here.