Elon Musk

by Ashlee Vance

  When speaking to potential recruits, Singh tried to energize them and be up front about the demands of SpaceX and of Musk at the same time. “The recruiting pitch was SpaceX is special forces,” she said. “If you want as hard as it gets, then great. If not, then you shouldn’t come here.” Once at SpaceX, the new employees found out very quickly if they were indeed up for the challenge. Many of them would quit within the first few months because of the ninety-plus-hour workweeks. Others quit because they could not handle just how direct Musk and the other executives were during meetings. “Elon doesn’t know about you and he hasn’t thought through whether or not something is going to hurt your feelings,” Singh said. “He just knows what the fuck he wants done. People who did not normalize to his communication style did not do well.”

  There’s an impression that SpaceX suffers from incredibly high turnover, and the company has without question churned through a fair number of bodies. Many of the key executives who helped start the company, however, have hung on for a decade or more. Among the rank-and-file engineers, most people stay on for at least five years to have their stock options vest and to see their projects through. This is typical behavior for any technology company. SpaceX and Musk also seem to inspire an unusual level of loyalty. Musk has managed to conjure up that Steve Jobs–like zeal among his troops. “His vision is so clear,” Singh said. “He almost hypnotizes you. He gives you the crazy eye, and it’s like, yes, we can get to Mars.” Take that a bit further and you arrive at a pleasure-pain, sadomasochistic vibe that comes with working for Musk. Numerous people interviewed for this book decried the work hours, Musk’s blunt style, and his sometimes ludicrous expectations. Yet almost every person—even those who had been fired—still worshipped Musk and talked about him in terms usually reserved for superheroes or deities.

  SpaceX’s original headquarters in El Segundo were not quite up to the company’s desired image as a place where the cool kids want to work. This is not a problem for SpaceX’s new facility in Hawthorne. The building’s address is 1 Rocket Road, and it has the Hawthorne Municipal Airport and several tooling and manufacturing companies as neighbors. While the SpaceX building resembles the others in size and shape, its all-white color makes it the obvious outlier. The structure looks like a gargantuan, rectangular glacier that’s been planted in the midst of a particularly soulless portion of Los Angeles County’s sprawl.

  Visitors to SpaceX have to walk past a security guard and through a small executive parking lot where Musk parks his black Model S, which flanks the building’s entryway. The front doors are reflective and hide what’s on the inside, which is more white. There are white walls in the foyer, a funky white table in the waiting area, and a white check-in desk with a pair of orchids sitting in white pots. After going through the registration process, guests are given a name badge and led into the main SpaceX office space. Musk’s cubicle—a supersize unit—sits to the right where he has a couple of celebratory Aviation Week magazine covers up on the wall, pictures of his boys, next to a huge flat-screen monitor, and various knickknacks on his desk, including a boomerang, some books, a bottle of wine, and a giant samurai sword named Lady Vivamus, which Musk received when he won the Heinlein Prize, an award given for big achievements in commercial space. Hundreds of other people work in cubicles amid the big, wide-open area, most of them executives, engineers, software developers, and salespeople tapping away on their computers. The conference rooms that surround their desks all have space-themed names like Apollo or Wernher von Braun and little nameplates that explain the label’s significance. The largest conference rooms have ultramodern chairs—high-backed, sleek red jobs that surround large glass tables—while panoramic photos of a Falcon 1 taking off from Kwaj or the Dragon capsule docking with the ISS hang on the walls in the background.

  Take away the rocket swag and the samurai sword and this central part of the SpaceX office looks just like what you might find at your run-of-the-mill Silicon Valley headquarters. The same thing cannot be said for what visitors encounter as they pass through a pair of double doors into the heart of the SpaceX factory.

  The 550,000-square-foot factory floor is difficult to process at first glance. It’s one continuous space with grayish epoxied floors, white walls, and white support columns. A small city’s worth of stuff—people, machines, noise—has been piled into this area. Just near the entryway, one of the Dragon capsules that has gone to the ISS and returned to Earth hangs from the ceiling with black burn marks running down its side. Just under the capsule on the ground are a pair of the twenty-five-foot-long landing legs built by SpaceX to let the Falcon rocket come to a gentle rest on the ground after a flight so it can be flown again. To the left side of this entryway area there’s a kitchen, and to the right side there’s the mission control room. It’s a closed-off area with expansive glass windows and fronted by wall-size screens for tracking a rocket’s progress. It has four rows of desks with about ten computers each for the mission control staff. Step a bit farther into the factory and there are a handful of industrial work areas separated from each other in the most informal of ways. In some spots there are blue lines on the floor to mark off an area and in other spots blue workbenches have been arranged in squares to cordon off the space. It’s a common sight to have one of the Merlin engines raised up in the middle of one of these work areas with a half dozen technicians wiring it up and tuning its bits and pieces.

  Just behind these workspaces is a glass-enclosed square big enough to fit two of the Dragon capsules. This is a clean room where people must wear lab coats and hairnets to fiddle with the capsules without contaminating them. About forty feet to the left, there are several Falcon 9 rockets lying next to each other horizontally that have been painted and await transport. There are some areas tucked in between all of this that have blue walls and appear to have been covered by fabric. These are top-secret zones where SpaceX might be working on a fanciful astronaut’s outfit or rocket part that it has to hide from visitors and employees not tied to the projects. There’s a large area off to the side where SpaceX builds all of its electronics, another area for creating specialized composite materials, and another for making the bus-sized fairings that wrap around the satellites. Hundreds of people move about at the same time through the factory—a mix of gritty technicians with tattoos and bandanas, and young, white-collar engineers. The sweaty smell of kids who have just come off the playground permeates the building and hints at its nonstop activity.

  Musk has left his personal touches throughout the factory. There are small things like the data center that has been bathed in blue lights to give it a sci-fi feel. The refrigerator-sized computers under the lights have been labeled with big block letters to make it look like they were made by Cyberdyne Systems, the fictional company from the Terminator movie franchise. Near the elevators, Musk has placed a glowing life-size Iron Man figure. Surely the factory’s most Muskian element is the office space that has been built smack-dab in its center. This is a three-story glass structure with meeting rooms and desks that rises up between various welding and construction areas. It looks and feels bizarre to have a see-through office inside this hive of industry. Musk, though, wanted his engineers to watch what was going on with the machines at all times and to make sure they had to walk through the factory and talk to the technicians on the way to their desks.

  The factory is a temple devoted to what SpaceX sees as its major weapon in the rocket-building game, in-house manufacturing. SpaceX manufactures between 80 percent and 90 percent of its rockets, engines, electronics, and other parts. It’s a strategy that flat-out dumbfounds SpaceX’s competitors, like United Launch Alliance, or ULA, which openly brags about depending on more than 1,200 suppliers to make its end products. (ULA, a partnership between Lockheed Martin and Boeing, sees itself as an engine of job creation rather than a model of inefficiency.)

  A typical aerospace company comes up with the list of parts that it needs for a launch system and then hands off their design an
d specifications to myriad third parties who then actually build the hardware. SpaceX tends to buy as little as possible to save money and because it sees depending on suppliers—especially foreign ones—as a weakness. This approach comes off as excessive at first blush. Companies have made things like radios and power distribution units for decades. Reinventing the wheel for every computer and machine on a rocket could introduce more chances for error and, in general, be a waste of time. But for SpaceX, the strategy works. In addition to building its own engines, rocket bodies, and capsules, SpaceX designs its own motherboards and circuits, sensors to detect vibrations, flight computers, and solar panels. Just by streamlining a radio, for instance, SpaceX’s engineers have found that they can reduce the weight of the device by about 20 percent. And the cost savings for a homemade radio are dramatic, dropping from between $50,000 to $100,000 for the industrial-grade equipment used by aerospace companies to $5,000 for SpaceX’s unit.

  It’s hard to believe these kinds of price differentials at first, but there are dozens if not hundreds of places where SpaceX has secured such savings. The equipment at SpaceX tends to be built out of readily available consumer electronics as opposed to “space grade” equipment used by others in the industry. SpaceX has had to work for years to prove to NASA that standard electronics have gotten good enough to compete with the more expensive, specialized gear trusted in years past. “Traditional aerospace has been doing things the same way for a very, very long time,” said Drew Eldeen, a former SpaceX engineer. “The biggest challenge was convincing NASA to give something new a try and building a paper trail that showed the parts were high enough quality.” To prove that it’s making the right choice to NASA and itself, SpaceX will sometimes load a rocket with both the standard equipment and prototypes of its own design for testing during flight. Engineers then compare the performance characteristics of the devices. Once a SpaceX design equals or outperforms the commercial products, it becomes the de facto hardware.

  There have also been numerous times when SpaceX has done pioneering work on advancing very complex hardware systems. A classic example of this is one of the factory’s weirder-looking contraptions, a two-story machine designed to perform what’s known as friction stir welding. The machine allows SpaceX to automate the welding process for massive sheets of metal like the ones that make up the bodies of the Falcon rockets. An arm takes one of the rocket’s body panels, lines it up against another body panel, and then joins them together with a weld that could run twenty feet or more. Aerospace companies typically try to avoid welds whenever possible because they create weaknesses in the metal, and that’s limited the size of metal sheets they can use and forced other design constraints. From the early days of SpaceX, Musk pushed the company to master friction stir welding, in which a spinning head is smashed at high speeds into the join between two pieces of metal in a bid to make their crystalline structures merge. It’s as if you heated two sheets of aluminum foil and then joined them by putting your thumb down on the seam and twisting the metal together. This type of welding tends to result in much stronger bonds than traditional welds. Companies had performed friction stir welding before but not on structures as large as a rocket’s body or to the degree to which SpaceX has used the technique. As a result of its trials and errors, SpaceX can now join large, thin sheets of metal and shave hundreds of pounds off the weight of the Falcon rockets, as it’s able to use lighter-weight alloys and avoid using rivets, fasteners, and other support structures. Musk’s competitors in the auto industry might soon need to do the same because SpaceX has transferred some of the equipment and techniques to Tesla. The hope is that Tesla will be able to make lighter, stronger cars.

  The technology has proven so valuable that SpaceX’s competitors have started to copy it and have tried to poach some of the company’s experts in the field. Blue Origin, Jeff Bezos’s secretive rocket company, has been particularly aggressive, hiring away Ray Miryekta, one of the world’s foremost friction stir welding experts and igniting a major rift with Musk. “Blue Origin does these surgical strikes on specialized talent* offering like double their salaries. I think it’s unnecessary and a bit rude,” Musk said. Within SpaceX, Blue Origin is mockingly referred to as BO and at one point the company created an e-mail filter to detect messages with “blue” and “origin” to block the poaching. The relationship between Musk and Bezos has soured, and they no longer chat about their shared ambition of getting to Mars. “I do think Bezos has an insatiable desire to be King Bezos,” Musk said. “He has a relentless work ethic and wants to kill everything in e-commerce. But he’s not the most fun guy, honestly.”*

  In the early days of SpaceX, Musk knew little about the machines and amount of grunt work that goes into making rockets. He rebuffed requests to buy specialized tooling equipment, until the engineers could explain in clear terms why they needed certain things and until experience taught him better. Musk also had yet to master some of the management techniques for which he would become both famous and to some degree infamous.

  Musk’s growth as a CEO and rocket expert occurred alongside SpaceX’s maturation as a company. At the start of the Falcon 1 journey, Musk was a forceful software executive trying to learn some basic things about a very different world. At Zip2 and PayPal, he felt comfortable standing up for his positions and directing teams of coders. At SpaceX, he had to pick things up on the job. Musk initially relied on textbooks to form the bulk of his rocketry knowledge. But as SpaceX hired one brilliant person after another, Musk realized he could tap into their stores of knowledge. He would trap an engineer in the SpaceX factory and set to work grilling him about a type of valve or specialized material. “I thought at first that he was challenging me to see if I knew my stuff,” said Kevin Brogan, one of the early engineers. “Then I realized he was trying to learn things. He would quiz you until he learned ninety percent of what you know.” People who have spent significant time with Musk will attest to his abilities to absorb incredible quantities of information with near-flawless recall. It’s one of his most impressive and intimidating skills and seems to work just as well in the present day as it did when he was a child vacuuming books into his brain. After a couple of years running SpaceX, Musk had turned into an aerospace expert on a level that few technology CEOs ever approach in their respective fields. “He was teaching us about the value of time, and we were teaching him about rocketry,” Brogan said.

  In regards to time, Musk may well set more aggressive delivery targets for very difficult-to-make products than any executive in history. Both his employees and the public have found this to be one of the more jarring aspects of Musk’s character. “Elon has always been optimistic,” Brogan said. “That’s the nice word. He can be a downright liar about when things need to get done. He will pick the most aggressive time schedule imaginable assuming everything goes right, and then accelerate it by assuming that everyone can work harder.”

  Musk has been pilloried by the press for setting and then missing product delivery dates. It’s one of the habits that got him in the most trouble as SpaceX and Tesla tried to bring their first products to market. Time and again, Musk found himself making a public appearance where he had to come up with a new batch of excuses for a delay. Reminded about the initial 2003 target date to fly the Falcon 1, Musk acted shocked. “Are you serious?” he said. “We said that? Okay, that’s ridiculous. I think I just didn’t know what the hell I was talking about. The only thing I had prior experience in was software, and, yeah, you can write a bunch of software and launch a website in a year. No problem. This isn’t like software. It doesn’t work that way with rockets.” Musk simply cannot help himself. He’s an optimist by nature, and it can feel like he makes calculations for how long it will take to do something based on the idea that things will progress without flaw at every step and that all the members of his team have Muskian abilities and work ethics. As Brogan joked, Musk might forecast how long a software project will take by timing the amount of seconds needed physicall
y to write a line of code and then extrapolating that out to match however many lines of code he expects the final piece of software to be. It’s an imperfect analogy but one that does not seem that far off from Musk’s worldview. “Everything he does is fast,” Brogan said. “He pees fast. It’s like a fire hose—three seconds and out. He’s authentically in a hurry.”

  Asked about his approach, Musk said,

  I certainly don’t try to set impossible goals. I think impossible goals are demotivating. You don’t want to tell people to go through a wall by banging their head against it. I don’t ever set intentionally impossible goals. But I’ve certainly always been optimistic on time frames. I’m trying to recalibrate to be a little more realistic.

  I don’t assume that it’s just like 100 of me or something like that. I mean, in the case of the early SpaceX days, it would have been just the lack of understanding of what it takes to develop a rocket. In that case I was off by, say, 200 percent. I think future programs might be off by anywhere from like 25 percent to 50 percent as opposed to 200 percent.