Elon Musk may be a pain, but that doesn’t mean his ideas are crazy

John Naughton


You don’t have to be a psychiatrist to wonder if Elon Musk, the founder of Tesla, is off his rocker. I mean to say, how many leaders of US public companies get into trouble with the US Securities and Exchange Commission for falsely claiming that they have secured funding to take their company private at $420 a share – and then get sued and fined $40m? Or can you imagine another CEO who deals with Wall Street analysts by swatting away questions about his company’s capital requirements as if they were flies. “Excuse me. Next. Next,” he replied to one guy who was pressing him on the subject. “Boring, bonehead questions are not cool. Next?”

The view from Wall Street is that Musk is too volatile to be in charge of a big and potentially important public company. The charitable view is less judgemental: it is that, while he may have a short fuse, he’s also a gifted, visionary disrupter. But even those who take this tolerant view were taken aback when he declared at a recent public event that he could see “one million robo-taxis on the roads by 2020”.

By robo-taxis he means self-driving Tesla cars. Given that Tesla has had such difficulty ramping up production of its Model S “affordable” sedans, the prediction sounded, well, nuts. But, on the grounds that one can never be sure with Musk, it seemed worth exploring further before calling for the men in white coats.

Some interesting facts emerge from even a cursory search. First, while Tesla doesn’t have the manufacturing skills of Ford or Toyota, it managed to deliver 90,000 vehicles in the final quarter of 2018. Second, it turns out that in the high-end US market, Tesla is outselling BMW, Mercedes and Lexus. And there are interesting developments in the European market too. One industry source claims that in February the Tesla Model 3 outsold the Mercedes C-Class, BMW 3-Series and the Audi A4. But this merely confirms that Tesla has finally learned to produce cars at volume, and that affluent customers are prepared to buy them.

Owners might be able to earn $30,000 in gross revenue a year from their cars

So what’s with this “robo-taxis” business? The key to it, according to Jean-Louis Gassée, a former Apple CEO and a seasoned observer of the tech industry, is that while Teslas might look like normal cars, in fact they represent a radically different conception of what an automobile should be. Although most cars nowadays have lots of electronics in them, their software is just an outsourced add-on to the hardware. In Teslas it’s the other way round: the hardware is the servant of the software. In those terms, the difference between an Audi A4, say, and the Tesla Model 3 is analogous to the difference between a Nokia feature-phone and an iPhone X.

What this means is that the nature and performance of a Tesla can be radically altered by a wireless software update. In 2017, for example, an update extended the range of Teslas in Florida to enable their owners to escape Hurricane Irma. But that’s small beer compared with the autopilot system that’s been available since 2014. Initially conceived as an advanced driver-assistance system, it has morphed into a system for a self-driving car. And it has been doing that in “shadow mode” by monitoring and logging every aspect of each Tesla’s journey – which means Tesla has terabytes of training data for its machine-learning systems.

What Musk foresees is a moment where Tesla owners can flick a switch that will turn their car into an autonomous vehicle that can function as a taxi (and therefore a revenue-earner) when it’s not needed. “The average Tesla car is parked for 22 hours per day,” he said. “Starting next year, owners will be able to flip a switch inside the Tesla app, and send out their car to pick up and drop off passengers autonomously, earning an estimated 65 cents per mile in fares.” Owners might be able to earn $30,000 in gross revenue a year from their cars, or more than $300,000 in revenue over the 11-year lifespan of their car.

This is an intriguing idea, but a lot has to happen on regulatory and other fronts before it becomes a reality – even in the US. It certainly won’t happen by 2020. Nor will there be a million Teslas on the road by then. Still, it’s not implausible in the longer term. Musk may be a pain in the ass, but he thinks out of the box. Besides, why should all the profits from self-driving cars go to the few corporations that will own and operate them?

What I’m reading

Talk really is cheap
“China only makes $8.46 from an iPhone – and that’s why Trump’s trade war is futile.” So runs the headline on a fascinating piece by Jason Dedrick, Greg Linden and Kenneth L Kraemer on the Conversation website.

The limits of consent
Go to Maciej Cegłowski’s Idle Words blog to read his bracing testimony to the US Senate on tech companies’ appropriation of our privacy.

That’s the way to do it

Read Hua Hsu’s lovely New Yorker essay on Umberto Eco’s classic manual How to Write a Thesis.