Another day, another runaway Tesla. This time it was a Model 3 in Irvine whose command-and-control functions unceremoniously shut down, leaving owner Javier Rodriguez stuck at 82 mph (133 kilometres an hour) on California’s busy Interstate 10. In this case, though the accelerator was unresponsive, Rodriguez’s brakes still worked, and he was eventually able, with help from the California Highway Patrol, to get off the freeway safely.
He was lucky. According to TeslaDeaths.com — yes, there really is such a site — there have been 12 Autopilot-related fatalities since Tesla introduced “self-driving” in 2015. Some, like Joshua Brown’s decapitation by semi-trailer, make headline news. Others, like the 2017 Model X that killed its 38-year-old driver by slamming itself into a lane divider on California’s famed Highway 101, get less attention. But Teslas Autopiloting themselves into collisions has become so common that Automotive News claims self-driving Model 3, S, X, and Ys have collided with no fewer than 12 first-responder vehicles, i.e. parked fire trucks attending to other accidents.
According to Bloomberg, the National Highway Traffic Safety Administration is finally started to take notice. Long accused of taking a “light touch” in regulating self-driving — and Tesla in particular — Bloomberg reports the administration is finally launching two formal defect investigations it says could ultimately force Tesla to retrofit cars with superior technology, and/or restrict the use of Autopilot.
It won’t come a moment too soon for the US National Transportation Safety Board, which, although it does not have the regulatory power the NHTSA wields, is in charge of accident investigations. And, according to Jennifer Homendy, the NTSB’s chair, “we essentially have the Wild West on our roads right now,” describing Tesla’s Autopilot — and its more recent Full Self Driving (FSD) Beta iteration — as “artificial-intelligence experiments” using untrained operators of 5,000-pound vehicles. She calls it all a “disaster waiting to happen.”
Politicians are starting to take notice, too. In fact, one California senatorial hopeful seems to be basing his entire electoral campaign on convincing the U.S. government to ban Tesla’s FSD. “The first danger I am tackling is Elon Musk’s reckless deployment of unsafe Full Self-Driving cars on our roads,” said Dan O’Dowd in announcing his campaign. “Tesla’s FSD software is the most unreliable safety-critical computer system I have ever seen.” His first advertisement is a full-blown one-minute public safety TV spot with a 1-855-STOP-FSD number, that runs through a compendium of Tesla self-driving close calls posted to YouTube.
Today I launched my campaign for U.S. Senate to make computers safe for humanity.The first danger I am tackling is @ElonMusk’s reckless deployment of unsafe @Tesla Full Self-Driving cars on our roads.
Visit my campaign website here: https://t.co/cBxWFjyhwS#ODowdForSenate pic.twitter.com/LOqslbKCPf— Dan O’Dowd (@RealDanODowd) April 19, 2022
Now to be sure, TV ads — most certainly, those of the political variety — are hardly bastions of veracity. And Dowd is, as the Teslarati are already clamouring, himself a tech CEO (of Green Hills Software), which may mean there’s a little inter-nerd rivalry behind his attack ads. But it doesn’t change the fact that, when your product becomes fodder for a political campaign — earlier this year, O’Dowd placed a full-page ad in The New York Times that assailed Tesla’s technology as “the worst software ever sold by a Fortune 500 company” — it just might be that your self-driving isn’t as “full” as you claim it to be.
Oh, and there’s one more startling fact before I (finally) get to the point of this story. According to Yahoo! News, other than the Volvo that killed poor Elaine Herzberg in Phoenix, Arizona — and that car’s automatic braking system, says Wired magazine, had been “overridden” by Uber — no one has been killed in a self-driving car from any other manufacturer. This despite the fact that virtually all automakers are now developing similar technology.
Cathie Wood, chief executive officer and chief investment officer, Ark Invest, gestures as she speaks during the Bitcoin 2022 Conference at Miami Beach Convention Center on April 7, 2022 in Miami, Florida. Photo by Marco Bello /Getty
All of which brings us to the latest prognostications by that most famous of Tesla boosters, Cathie Wood. In her latest ARK Invest proclamation, Wood, a mega-loyal Tesla booster, says that TSLA’s market capitalization will grow to US$5.3 trillion by 2026. It now stands at a little over US$1 trillion, which would require a more than 400-per-cent increase in just four years. Nothing spectacular about that, you say, as Wood is always pumping — and sometimes “dumping” — TSLA shares.
Nonetheless, this latest missive does stand out as particularly fantastical because Wood claims the lion’s share — some 62 per cent, in fact — of that future market cap will be Tesla’s yet-to-be-finalized robotaxi business. That, with just a flick of the old calculator, works out to be US$3.3 trillion-with-a-t dollars. And, doing a little more rudimentary math, that also means TSLA’s robotaxi business will be valued about US$3,200 a share all by its own in just four years. Oh, and by the way, Elon Musk just announced that Tesla won’t produce its first robotaxi — without steering wheel or pedals — until 2024.
Spacex founder Elon Musk celebrates after the successful launch of the SpaceX Falcon 9 rocket with the manned Crew Dragon spacecraft at the Kennedy Space Center on May 30, 2020 in Cape Canaveral, Florida. Photo by Joe Raedle /Getty
Speaking of which, in 2026 — just four years hence, I’ll again remind you — Wood also estimates that said revenue will somehow grow to US$287 billion. Even by the standards of incredulous, that simply beggars belief. First off, the company whose name — Uber — has become synonymous with ride-hailing — as in “I’ll be Ubering over soon” — had, according to Statista.com, a net revenue of US$17.46 billion in 2021. Globally.
That means — and now I need a bigger calculator — that Wood expects Tesla’s robotaxi business to grow to 15-plus times the size of Uber in just four years. And just so we’re clear, she’s not talking about human-driven ride-hailing, which she carves out as another, albeit much smaller, five-per-cent slice of Tesla’s 2026 revenues. That’s US$42 billion, which means that Wood sees just the sideline of Tesla’s ride-hailing business — the human-driven portion — overwhelming Uber by some 240 per cent in four years, even though Tesla won’t actually produce any of those pedal-less robotaxis for another two years. Please feel free to giggle now.
When your product becomes fodder for a political campaign, it just might be that your self-driving isn’t as “full” as you claim it to be
That’s not to say all of Wood’s predictions are doomed. Even if ARKK’s revenue projections are ludicrous, it’s still possible her TSLA share price targets could be met. However, revenues and TSLA share prices have not traditionally been well-acquainted, Tesla’s power amongst retail investors making the GameStop short-squeeze look like dime-store shopping.
Nonetheless, all this posturing will have consequences. As this discussion becomes ever more polarized — Wood claiming that Tesla will essentially become a taxi company in four years, while O’Dowd’s extremely well-funded political campaign tries to get FSD ‘cancelled’ — one hopes it doesn’t spur more “Look, ma, now I’m in the back seat” Autopiloting. Nonetheless, experience suggests that the more Tesla’s FSD is condemned as inferior, the more Musk cultists might try to self-drive their Model Xes to even more extreme trials. In the meantime, we wait for the regulators to enact directives so that everyone — consumers and automakers alike — will know what the rules are. Laws again that they — at least the NHTSA — seem loathe to craft or prosecute.
A Tesla crashed into a fire truck near Los Angeles in 2018. Photo by Culver City Firefighters
Like all such regulations, the devil will be in the details. Or, more accurately, in the definitions. The Law Commission for England and Wales and the Scottish Law Commission, for instance, recently unveiled a proposed Automated Vehicles Act which would allow cars to drive themselves without “human monitoring.” To be classified as “self-driving,” the onboard computers would have to be able to prove they could “control the vehicle so as to drive safely and legally, even if an individual is not monitoring the driving environment, the vehicle or the way that it drives.”
The problem, according to AutomotiveWorld.com, is that the entire bill — and ostensibly other bills like it — hangs on exactly what they mean by “drive safely.” If, as AW suggests, that turns out to mean as safe as “a competent and careful human driver,” then it’s a threshold easily reached. If autonomous automobiles are required to be much safer — or, as much of the general public still believes, completely without fault — then widespread Level 5 self-driving may never become a reality. Which means that Tesla’s robotaxi business will never generate anything like US$287 billion in net revenues.
That requisite standard of safety will ultimately, the commissioners suggest, become a political question. In the end, that means we consumers — a.k.a. the voting public — will become the final arbiters of what will and won’t be acceptable autonomous driving. If we demand that robotic cars need only be safer than human drivers, fully autonomous automobiles will surely become a reality soon. On the other hand, if we as a society continue, as I suspect, to be less forgiving of mistakes caused by robots — and the fatalities that result — the chances of autonomously driven ride-hailing taxis becoming popular is not good at all. Ditto, of course, Cathie Wood’s prognostications.
Keyword: Motor Mouth: Critics could cancel Tesla's robotaxi before it gets made