Over the last two years, Tesla has been charging customers $3,000 to $5,000 for an upgrade called “full self-driving.”
“All you will need to do is get in and tell your car where to go,” Tesla’s order page said. “Your Tesla will figure out the optimal route, navigate urban streets (even without lane markings), manage complex intersections with traffic lights, stop signs and roundabouts, and handle densely packed freeways with cars moving at high speed.
Last week, Tesla abruptly dropped the option from the order page. It was a tacit admission that the technology was not close to being released. But it’s been obvious for months that Tesla has been struggling to deliver on its self-driving promises.
In his October 2016 conference call announcing Tesla’s second-generation autopilot platform, Tesla CEO Elon Musk claimed that “all cars exiting the factory have hardware necessary for Level 5 Autonomy”—industry jargon for cars that can drive without human intervention on all roads and in all weather conditions. Musk predicted that a Tesla vehicle would be able to drive from Los Angeles to New York without human intervention by the end of 2017.
Needless to say, Tesla didn’t demonstrate the ability to do a coast-to-coast driverless trip in 2017. And it’s now clear that Tesla cars manufactured since October 2016 don’t actually have the hardware necessary for full self-driving capability as Musk claimed—Tesla announced upgraded silicon back in August that will be required for full self-driving capabilities. Musk has said that customers who paid for the full self-driving feature will get a free upgrade.
So how do customers who already paid thousands of dollars for the full self-driving technology—technology that still seems to be years away, if it ever becomes available—feel about the news that Tesla has halted orders? When I put out a call on Twitter on the subject, I was expecting to hear from a lot of angry and frustrated customers. But overall the results were surprisingly positive. I talked to six Tesla customers on the phone and heard from a few others via email. Only one of these had negative things to say about the way Tesla sold the full self-driving feature.
Most of the happy customers I talked to seemed to be clear-eyed about the risk they were taking—they knew that full self-driving technology was likely to be years away.
Elon Musk “usually always delivers, but his deadlines aren’t always the most accurate,” said Charles Scott, who lives in the Houston Area and bought a Model 3 this summer. When it comes to delivering on his promises, Scott says, Musk operates on “Elon time.”
Scott says he knows full well that he’s “not going to be taking a nap in the back of the car in the next 6 months.” He expects full self-driving capabilities to come out in two to three years. And he expects that those who pre-paid for the full self-driving capabilities will get any necessary hardware upgrades for free.
It’s a sentiment echoed by other Tesla customers I talked to. One customer told me that she wouldn’t be surprised if it took a decade for Tesla to achieve full self-driving.
“There’s this idea in the media of Tesla scamming people out of money,” said Model 3 owner Earl Banning. “I feel like it was the opposite experience.”
Banning initially turned down the full self-driving option, then later changed his mind and asked to take the upgrade. He said he felt that he was under no pressure to take the upgrade; on the contrary, it took some persistence to get Tesla to sell him the feature. That’s partly because Tesla was in the process of raising the option’s price—Banning was grandfathered into the old price but Tesla’s website wasn’t set up for that. But Banning also says that Tesla reps made sure he knew that the option wasn’t actually functional yet.
Banning, Scott, and other customers expect Tesla to roll out a steady succession of enhancements to Autopilot functionality, perhaps better handling situations like stoplights, on- and off-ramps, and so forth. As long as Tesla continues to improve the cars’ driving abilities, they’ll feel like they’re getting their money’s worth, even if it takes many years to get to full self-driving capabilities.
“It was sort of like getting an early ticket to the show,” Scott said. It’s worth it for “the entertainment value.”
The five happy customers I talked to were all Model 3 owners who bought their cars in the last six months. But I also talked to a man in Birmingham, Alabama, who asked that we not use his name in this story. He bought a Model S in October 2016, days after Tesla first announced the full self-driving feature. And he had a very different perspective.
“I was expecting to have it in about 6 months,” he said about full self-driving capabilities. “Two years later, I’m more aware of Elon’s propensity to oversell things.”
Two years ago, he wasn’t a Tesla news junky, and he took Tesla’s optimistic marketing at face value. And he feels like he got burned.
You shouldn’t read too much into the exact ratio of positive and negative responses here. All the customers I talked to responded to a tweet I posted on Tuesday evening. That’s the opposite of a random sample, and it’s quite possible that passionate Tesla fans are over-represented (or under-represented) in my corner of Twitter. But it’s clear that there are a lot of Tesla customers who are happy to give Musk more time to deliver on his full self-driving promises—as well as some who aren’t so happy about the situation.
Tesla may be many years away from full self-driving
While some Tesla customers remain optimistic that Tesla will eventually achieve full self-driving capabilities on today’s Tesla cars, a lot of experts and industry insiders believe that the problem is much harder than Musk is willing to admit.
Probably the most important factor is that Tesla cars don’t have lidar. Every other major company working on self-driving technology has cars festooned with lidar sensors. Lidar is valuable because it provides a highly accurate three-dimensional map of the surroundings, greatly simplifying the difficult task of understanding the car’s environment.
No one doubts that it’s theoretically possible to navigate without lidar. After all, human beings do it with a pair of cameras called eyes. But there’s good reason to think that cars with lidar will be able to navigate autonomous long before cars without lidar can.
Tesla’s current computing hardware is also likely to be underpowered for full self-driving capabilities—a point Tesla tacitly conceded when it announced that it will be rolling out a new, far more powerful chip next year. Musk has said current customers who paid for the full self-driving option will get a free upgrade.
The other issue is that Tesla doesn’t seem to be laying the kind of groundwork that’s required for fully autonomous vehicles. Competitors like Waymo and Cruise have built detailed maps of areas where they’ll be operating, and they have large fleets of cars with safety drivers testing out the capabilities of their software. Tesla, by contrast, told California regulators earlier this year that it hadn’t logged a single mile of public road testing in California.
It’s certainly possible that Tesla has figured out a way to achieve full self-driving capabilities that don’t depend on extensive real-world testing—maybe data collected from existing customers can be a substitute for a purpose-built fleet of test vehicles. But Tesla certainly hasn’t explained how this will work, and there’s a risk that the company will hit a dead-end trying to validate its vehicles without conventional on-road testing.
Tesla isn’t communicating clearly about full self-driving
While Tesla has pulled the full self-driving feature from its order page, it hasn’t stopped touting the capability on its website. The official Tesla Autopilot page still has a banner that advertises “full self-driving hardware on all cars” across the top—despite the fact that it’s widely believed that full self-driving will require upgraded silicon Tesla plans to roll out next year. The page has the same marketing copy about full self-driving that used to be on the order page, promising customers that “all you will need to do is get in and tell your car where to go.”
The page does say that full self-driving is “dependent upon extensive software validation and regulatory approval” and that it is “not possible to know exactly when each element of the functionality described above will be available, as this is highly dependent on local regulatory approval.”
That strongly implies that the technology actually exists and just hasn’t been approved by regulators. But it would be surprising if this were true, since Elon Musk isn’t known for being shy about touting the technological breakthroughs of his companies. It seems more likely that the timing of the features is uncertain because Tesla hasn’t created them yet. We’ve asked Tesla for comment and will update if the company replies.
In a tweet last week, Musk said that Tesla removed the full self-driving option from the order page because it was “causing too much confusion.” But removing it from the order page creates confusion, too. And the fact that full self-driving is still prominently featured on Tesla’s Autopilot page without any mention of a separate charge could easily cause new customers to think that full self-driving will be a free upgrade for those who purchase the Autopilot option.
Tesla benefits from intensely loyal fans
Tesla customer Charles Scott compares the full self-driving option to a crowdfunding campaign. It’s a good analogy—and one that was endorsed by other Tesla customers I talked to. People who contribute to a Kickstarter campaign typically expect to get something in return. But backing a Kickstarter project isn’t like a normal product purchase. Customers are usually excited about the concept behind a project and want it to succeed. They understand and accept that the project has risks and that the product may be delivered late—or not at all.
It’s hard to imagine most other big companies employing this kind of strategy. Customers would be livid if an ordinary technology or car company charged customers for a new product or feature and then failed to deliver it for two years or more.
But Tesla isn’t a normal company. Tesla customers seem to be intensely loyal to the company and its CEO. Many see themselves as joining a movement for a more sustainable car industry as much as they’re buying a consumer product. They feel invested in Tesla’s long-term success, and, as a result, they’re willing to cut Tesla a lot of slack when the company blows through deadlines or asks customers to take a leap of faith.
“I believe in the company,” said Tesla owner and full self-driving purchaser Earl Banning. “I believe in Tesla’s mission.”
Even the Alabama Model S owner who told me he felt ripped off by the full self-driving upgrade said he’s otherwise very happy with his car.
And this seems like a reason to be bullish about Tesla regardless of how the company’s full self-driving project ultimately turns out. While some customers are angry at Musk for overhyping the self-driving feature, many others are happy to be patient and see what Tesla comes up with in the coming years. They know that Elon Musk often fails to deliver on his promises—or at least on his ambitious self-imposed deadlines—but they love Tesla and its CEO anyway. And those Tesla customers’ extreme willingness to cut Musk slack gives him the freedom to take on ambitious projects that would seem too risky for most CEOs.