Driverless cars aren’t going away, but we need to lower our expectations about them

/

The failure of Argo AI may lead some to speculate that autonomous vehicles are a lost cause. But the technology works, often very well; it just won’t be used in the way we originally thought.

Driverless car surrounded by regular cars

Driverless cars aren’t going to change the world.
Illustration by Alex Castro / The Verge

The driverless car startups that haven’t run out of money, gone out of business, or been bought out want you to know that driverless cars aren’t going away. These companies insist that skepticism about the technology is expected but unwarranted and that driverless cars will, in the words of one co-founder, “change the world.”

Where have we heard this before? Oh right, practically every tech company and Silicon Valley partisan that has emerged in the last decade or so has insisted that their only goal is to empower mankind and change the world (and not, I don’t know, make gobs and gobs of money).

I do agree autonomous vehicles are coming (many are already here), but I think we need to rein in the “change the world” rhetoric. Like, significantly. Let me explain why.

Earlier this week, an AV operator shut down. This was unexpected for many reasons. Argo AI had the financial support of two major automakers: Ford and Volkswagen. The company was testing its vehicles in Miami, Austin, and Washington, DC — major markets that were expected to become launchpads for Argo’s eventual robotaxis and self-driving trucks. And the company was considered one of the good ones, in so far as its leaders were taking things like safety and edge cases very seriously.

But even though it appeared to be doing everything right from a technology and regulatory perspective and had the backing of two very powerful car companies, Argo still fell victim to the growing awareness that driverless cars were years, if not decades, away from being safe and reliable enough to prop up a sustainable business. Ford said as much during its earnings report this week, admitting that “fully autonomous vehicles at scale are a long way off.” Better to cut ties now than remain continually on the hook for years to come.

In retrospect, maybe we should have been more skeptical about headlines like this. And “expert” analysis like this. And guys who predict things like, “We will have more than one million robotaxis on the road.”

From a business perspective, it’s not surprising to see companies like Argo fail, said Raj Rajkumar, professor of engineering and robotics at Carnegie Mellon University. “Unfortunately, the basic business fact is that these robotaxi companies have been spending tons of money with zero revenues,” he said in an email to The Verge. “In business, this situation simply cannot last long. Argo AI is the first domino to fall — expect others in the robotaxi sector to follow suit.”

This has been clear for a while now. Argo is certainly the most substantial company to call it quits, but there have been a growing number of acquisitions, sell-offs, and bad business moves in the AV space for years.

Zoox sold out to Amazon. Uber practically gave away its AV division for free to Aurora. Lyft sold to a subsidiary of Toyota. Cruise bought Voyage. Nuro acquired Ike. (I assure you, you’re not having a stroke — these are just the quirky names of various AV startups.)

The companies that are still around are hemorrhaging money. Aurora, which absorbed Uber’s discarded division, is said to be mulling a sale to Apple or Microsoft. The company went public last year by merging with a special purpose acquisition company (SPAC), and then lost about 80 percent of its value. This is the same company that was started by Chris Urmson, one of the founders of the Google self-driving car project (now Waymo), a guy once called the “Henry Ford of autonomous driving,” who said he hoped his kids will never have to get driver’s licenses.

One day, you think you’re going to change the world; the next, you’re watching your stock price plummet and weighing whether to beg Tim Cook for a bailout.

But when I say that driverless cars aren’t going away, I mean that the technology is likely here to stay. And that’s because it works, not always reliably and sometimes a bit trepidatiously, but it does work. I’ve been in the cars. I’ve interviewed the engineers. I’ve watched this technology evolve and grow for the last seven years.

What is changing is the hype. The big bets aren’t paying off. Automakers and big venture capital firms that were bullish about the prospect of millions of robot cars roaming the streets in the near future are losing money. Like lots of money. SoftBank, which poured billions of dollars into AV projects like Cruise, Nuro, and Uber, recently reported losing $23 billion from its Vision Fund investments.

I think Ed Niedermeyer, co-host of the podcast Autonocast, got it right when he tweeted, “We make everything all about technology these days, the good and the bad alike, but the ‘self-driving car bubble’ was a manifestly human failure. Hubris, greed, manipulation and self-deception are what failed in the AV sector… not technology.”

I mean, we’ve gotten to the point where AV operators have openly acknowledged that driverless cars appear to be stuck in the “trough of disillusionment,” a reference to Gartner’s Hype Cycle that tracks the maturity of various technologies. 

During General Motors’ third quarter earnings call this week, Cruise CEO Kyle Vogt said, “As for the industry, we’re seeing increased separation between the company’s operating commercial driverless services and those that are still stuck in the trough of disillusionment. What’s happening here is that the companies with the best product have pulled ahead and are accelerating.”

Cruise is aiming for $1 billion in revenue by 2025. This past quarter, it lost $497 million for a total deficit of $1.4 billion so far this year. It also says it will launch commercial services in Phoenix and Austin before the end of the year. And Waymo is bringing its robotaxis to Los Angeles. The slow, steady march continues.

So, if we’re not going to have robot cars that can magically whisk us wherever we want, what are we left with? There’s a lot of excitement around autonomous construction and farm vehicles. Autonomous warehouse equipment seems likely. Robot semi trucks that can relieve a lot of the burdens of long-distance deliveries remain viable. Autonomous trains and buses are also a strong possibility, especially when they operate on a fixed track without a lot of complexity.

The one thing that all of these applications have in common is that they are difficult to hype and therefore less likely to receive pallets of cash in the same way that robotaxi startups became magnets for VC enthusiasm. Driver-assist technology, like Ford’s BlueCruise, GM’s Super Cruise, and Tesla’s Autopilot, is likely positioned to soak up some of the excess hype as robotaxis continue to trend downward. But that technology is not without its pitfalls, too.

So where does that leave us? Pretty much in the same place we’ve been. Cars remain a dominant force in our lives, for better or worse (mostly for worse), and robotaxis were always an unlikely candidate to “change the world.” Maybe you won’t ever own your own driverless car or even ride in one like an Uber. But you might drive past a construction site using autonomous backhoe loaders or take a train without a conductor.

There are still possibilities. They just aren’t endless.

Go to Source