Ludicrous: The Unvarnished Story of Tesla Motors

by Edward Niedermeyer

Tesla disclosed that “only four percent of customers chose the 40 kWh battery pack, which is not enough to justify production of that version.” But because customers could only order vehicles then in production, at least until Tesla switched its order process at the beginning of 2013, it’s not clear that the vast majority of Model S reservation holders had even been able to register their preference. The 40 kWh version was also never capable of using Superchargers, eliminating one of Tesla’s most critical advantages.

Nobody even seemed to care that Tesla’s tiny one-time profit didn’t show any real potential for recurring profits, or that its promises for the future couldn’t hold up to even the most cursory scrutiny.

At a time when making money in the online media business was becoming increasingly difficult, the twin inducements of stock promotion and affiliate marketing had a profound effect on perceptions of Tesla.

But just as Tesla itself was sucked into a cycle of overpromising that eventually left it totally dependent on ever-increasing hype, so too did its fan base find itself shifting from purely passionate advocates to investors who could boost a highly lucrative stock and earn rewards while promoting the company they loved and helping it sell cars.

If you look at some of the unicorns with business models that I just cannot make work, investors are throwing money at these companies,” he said, naming the commercial real estate company WeWork as another example.

Common arguments against EVs are that they tie drivers to a much shorter range than conventional vehicles, and that driving one doesn’t actually cut down your carbon footprint—because an electric car has a figurative “long tailpipe” that shifts the source of emissions from the vehicle itself to the power plant.

Musk’s most important promise of the night—that every Supercharger would not only be solar powered but in fact generate more solar power than Tesla drivers would use—never came close to materializing.

A Supercharger station with ten stalls, each providing an average of ten 50 kWh charges per day, would require solar arrays capable of generating 5 mWh per day. Even in the most consistently sunny parts of the world, where solar panels can generate about 1 kWh per square meter, more than an acre of panels would be required to even come close to this level of output.

Most studies conclude that electric cars powered by even relatively dirty mixes of coal and natural gas pollute less than gasoline cars, so any Tesla should be an improvement on the status quo. Indeed, this was the core insight that led Eberhard and Tarpenning to create Tesla in the first place.

When operating at full potential, a Supercharger still takes half an hour to provide an 80 percent charge, with a full charge taking more than an hour. Practically speaking, this translates to between two and three hours of driving for every hour spent charging.

Though this may be the case for well-off vacationers traveling at a leisurely pace, it simply can’t be compared to the five minutes it takes to refuel a car with gasoline.

Overnight, Tesla’s 85 kWh Model S went from earning four credits per vehicle to seven. Moreover, to earn this dramatic increase in credits, Tesla needed only to prove to CARB that such rapid refueling events were possible. By demonstrating battery swap on just one vehicle, Tesla nearly doubled the ZEV credits earned by its entire fleet even if none of them actually used the swap capability.

Tesla’s cynical use of battery swap echoed its claims to be taking huge amounts of carbon off the road while failing to deliver on its solar power promises and even using diesel generators at Harris Ranch.

This pattern of repeatedly promising environmental benefits and then not fully delivering on them but claiming the benefits for publicity purposes suggests that Tesla’s use of environmentalism is less purely principled than Musk claims.

Tesla still had yet to provide proof that even a single Supercharger was actually fully solar-powered . . . and yet new promises of off-grid Superchargers were still coming. Whether these promises are noble aspirations or cynical lies almost doesn’t matter at this point.

But ultimately the Brown crash was not so much about a technical failure on the part of Autopilot, but the deadly risk Tesla was running by marketing an ADAS system as a nearly autonomous car.

But there was far more to Musk’s statistical defense of Autopilot’s safety than just “doing the bloody math.” Musk was comparing a tiny number of Teslas, none of which was more than four years old or cheaper than about $70,000, to a much larger US car fleet with an average age of eleven years and an average price of $35,000. Moreover, Autopilot was supposed to be used only on divided highways, whereas the average car operates in far more complex environments, making crashes more likely.

“So if you don’t brake, it’s your fault because you weren’t paying attention,” she explained. “And if you do brake, it’s your fault because you were driving.”

As Tesla’s own supplier of Autopilot-enabling hardware and others pointed out, Tesla’s decision to rush Autopilot to market, push the technology to the limit, and overhype its capabilities was what put lives—and the future of autonomous-drive technology itself—at risk.

but Tesla’s record compares poorly to that of the Cadillac Super Cruise system, which does exactly what NTSB called for: limits use to mapped freeways and tracks the driver’s gaze to prevent distraction. Not a single crash has been reported in a vehicle with Super Cruise activated.

what people actually want out of them are both the ability to pay less attention when behind the wheel and the high-tech cool factor of a car appearing to drive itself.

From its earliest days, the auto industry had been dogged by a persistent challenge: smaller, cheaper cars are no less expensive to develop—and nearly as expensive to build—as large, expensive cars.

But the reason established automakers don’t offer flashy features like falcon-wing doors is not a lack of design talent or creativity, as evidenced by decades of concept cars featuring outrageous designs that rival anything Tesla has produced. Rather, the reason production cars never look as flashy as the design concepts you see at car shows is that automakers have all learned the lesson that Tesla was about to learn with the Model X.

Many of the early posts described problems similar to the ones that plagued the Model S, including poor paint quality, door and window seal problems, interior squeaks and rattles, misaligned body panels and trim, and generally poor fit and finish. But the early Model Xs also had a host of unique problems, including faulty air-conditioning compressors and weird double- and triple-vision illusions from the windshield in low light conditions.

The complexity of the Model X’s front and rear falcon wing caused countless problems:

Because the front and rear door handles were designed to meet up, even relatively minor misalignments were instantly identifiable since the handles didn’t quite mirror each other.

There was simply no getting around the fact that Musk’s design ambitions had radically exceeded Tesla’s manufacturing capabilities.

Any car company that must rely on the physical presence of its CEO to manufacture vehicles to an acceptable quality standard is essentially admitting that its culture isn’t committed to quality.

The most brilliant design in the world is worth very little if it can’t be made efficiently, profitably, and at competitive quality. These “compromises” aren’t the products of generic corporate mediocrity, but of survival of the fittest

but in general the Model X served only to deepen Tesla’s fan base at a time when it desperately needed to broaden it.

The scale and complexity of modern automobiles has made the car business the ultimate team sport, relegating all-powerful automotive auteurs to either tiny luxury-brand workshops or the history books.

As a result of this cultural vacuum, almost every decision that management makes at Tesla ends up coming down to a single consideration: Will Elon approve?

The agreement stated that in exchange for goodwill repairs, the customer releases Tesla from any claims or damages, promises not to sue Tesla for anything related to the problem being fixed under goodwill, and agrees to keep the entire incident confidential.

Even if no underlying defect were ever provable, the mere existence of an NDA that could prevent an owner from reporting a defect to NHTSA was a serious safety concern.

“There is no car company in the world that cares more about safety than Tesla,” the post argued, as if examples of its compliance with defect-reporting laws were somehow extraordinary and not the minimum legal behavior for a company selling new cars.

By reaching out to owners, Tesla was effectively performing a “stealth recall” of a part it knew could cause vehicles to stall dangerously.

Even after the events surrounding its high-voltage contactors, and the recalls of its drivetrains in Toyota and Daimler products, Tesla proudly noted that “none of our past recalls or corrective measures have been related to our electric powertrain.”

Tesla’s problems on safety come down to some of the unique attributes that it brings from the software world into the highly regulated, life-and-death world of automobiles.

its approach of launching with a “minimum viable product,” and then iterating the vehicle’s engineering over time not only makes it more likely to bring an undertested and potentially problematic vehicle or component to market, it also means regulators can’t properly track vehicle populations to oversee safety. Tesla’s seemingly systematic evasion of NHTSA’s independent investigatory capabilities also brings to mind Silicon Valley’s cultural disregard for fussy regulations and slow-moving bureaucrats.

Without a solid business to support its heady valuation, a loss of confidence could have catastrophic, cascading consequences that could affect its ability to hire talent, raise fresh capital, and ultimately survive.

But Musk was trying to fit a square peg into a round hole: a car designed for private ownership would never be optimal for taxi duty, particularly not the fast but fragile premium cars Tesla built.

Tesla had the dubious distinction of being the first maker of automated driving technology to have one of its owners die in a crash while the system was activated, and the NTSB’s investigation clearly identified design decisions that had contributed to Josh Brown’s death.

These numbers suggested that Tesla’s system performed far less effectively compared to Waymo’s average of more than 5,000 miles between disengagements and BMW’s single disengagement in 638 miles of testing. It also proved that Tesla was testing its “self-driving” system far less than its competitors: just 550 miles, compared to Google’s 635,868 miles, the automotive supplier Delphi’s 16,662 miles, and VW’s 14,945 miles in the same year.

The fact that the vast majority of Tesla’s “test” miles came while it was filming its demonstration video only added to the impression that Tesla was spending more time hyping its newest unfulfilled promise than it was working toward delivering on it.

autonomous-drive technology experts were almost unanimously skeptical of Tesla’s ability to deliver on Musk’s promises, let alone challenge Waymo for leadership in the space.

Tesla’s sensor suite only had redundant and diverse sensors in the front of the car where its radar and cameras overlapped, whereas Waymo and others had 360 degrees of redundant sensor coverage from cameras, radar, and lidar.

Each sensor type has advantages in different tasks. For example, radar and lidar excel at determining an object’s range, while cameras are generally superior at recognizing and classifying objects.

The fundamental challenge of driving is that most of it is boring and predictable routine, punctuated by the sudden appearance of unexpected situations—an animal crossing the road, a bale of hay falling out of the truck in front of you, a distracted driver plowing through a red light—that require skill, experience, and presence of mind to safely navigate.

Silicon Valley’s “move fast and break stuff” ethos, which fostered innovation in non-safety-critical software, was beginning to look like a poor approach to developing self-driving car systems, which hold the power of life and death.

Any plan to simply update software whenever an underbaked autonomous-drive system runs into a situation it can’t handle has to assume that a number of those situations will involve a customer dying.