Just before President-elect Donald Trump met with leaders from Apple, I.B.M., Amazon, Facebook, and other tech companies in mid-December, the transition team made a somewhat puzzling announcement: Elon Musk, the head of Tesla, SpaceX and SolarCity, who would be attending the session in Trump Tower, had been named to the President’s Strategic and Policy Forum, a new committee intended to advise the President on economic policy. The appointment seemed odd chiefly because Musk’s political views-he has been pro-immigration and against the Keystone XL Pipeline while supporting a carbon tax on businesses and expressing passionate concern about climate change-would seem to veer far from Trump and his party. Still, whatever their ideological differences, Trump and Musk are not such an unlikely duo; they have many qualities in common. Both are outsized businessmen and public figures who have enjoyed hoopla and credulous press for their projects despite checkered records of success. Musk is best known as an innovator with unorthodox ideas. He created a car company from the ground up, building luxury electric cars with a devoted following. He jumped into solar energy and battery development on a big scale, vows to colonize Mars, and has said that we are probably living in a computer-simulated world. But, while his grand gestures inspire awe and curiosity, they often fall short in the execution. Since 2011, Tesla has failed to meet Musk’s product-launch, production, and financial-performance promises more than twenty times, according to an analysis by the Wall Street Journal. Even a private showing, in early January, of Tesla’s new Gigafactory, in Storey County, Nevada-which Musk claims is on schedule to mass-produce lithium-ion batteries at rock-bottom costs by 2018-didn’t instill confidence in Musk’s ability to achieve his stated goals. As the Pacific Crest Securities research analyst Brad Erickson said in a note, the tour left “much to the imagination.” And in September an explosion destroyed an unmanned SpaceX rocket on the launch pad during a fuelling exercise-an incident that called into question the viability of Musk’s radical notion to refuel craft en route, with astronauts on board. A little more than a year earlier, a NASA-funded SpaceX rocket carrying cargo destined for the International Space Station exploded two minutes after lift-off, destroying the payload. A NASA report on that incident raised questions about quality standards at Musk’s company. The first unmanned SpaceX flight since the September accident is scheduled for later this month. Musk, however, is nothing if not self-assured, a quality he brings to an upcoming task: he has promised to devise an entirely new system for making automobiles, with levels of efficiency and productivity that have never been achieved before. He intends to do this with something he calls a “machine that builds the machine,” a nearly fully automated factory that will assemble cars at speeds greater than any other global automaker. Beginning in 2018, this plant, an extension of Tesla’s current facility in Fremont, California, is slated to produce as many as half a million vehicles each year, and Musk has previously promised to increase output by some fifty per cent every year. This is an extraordinary leap in production: in 2016, Tesla produced fewer than eighty-four thousand cars. In the auto industry, Musk’s production assertions are viewed as the manufacturing equivalent of vaporware-an advance that is promised but has very little chance of becoming a reality. The most efficient automakers-companies such as Toyota and Honda that have spent decades studying, designing, and perfecting manufacturing systems and processes-are unable to approach anywhere near the levels of production that Tesla, which isn’t fifteen years old, is aiming for. Indeed, a prolific auto plant in the U.S., running three six-and-a-half-hour shifts a day, produces just more than four hundred thousand cars a year-and only if it operates at peak efficiency virtually all of the time, without interruptions to deal with parts shortages, incorrectly sized or mismatched components, or design changes. The disparity between real-world factory output and Musk’s goals likely means that many of the nearly four hundred thousand people who have already put down a thousand dollars to reserve a Tesla Model 3, which Tesla says will cost about thirty-five thousand dollars, will have to wait until 2019 or later for their cars. Even one of Tesla’s most stalwart supporters in the investment community, the Morgan Stanley analyst Adam Jonas, has reached the conclusion that the Model 3s will be late, and that there will be fewer of them than advertised. In November, he wrote, “We continue to forecast a Model 3 launch at the very end of 2018 (more than 1 year later than company target) with 60k units in 2019 and 130k units in 2020.” Musk does not give many specifics about how he will produce so many cars, so quickly, in ways that have eluded his more experienced competitors. But he is unambiguous about one foundational principle: his ramp-up in production can come about only by eliminating humans from the manufacturing process as much as possible. “You really can’t have people in the production line itself, otherwise you drop to people speed,” Musk said on an earnings call in August. In this context, it’s worth noting that the aim of the Trump advisory committee that Musk was appointed to is to figure out ways to “make it attractive for firms to create new jobs.” Musk’s plan for a fully automated, so-called “lights-out factory”-because robots can work in the dark-is a strategy many in the industry consider to be a short-sighted. Humans can help improve efficiency by identifying shortcomings and conceiving of ways to do things that will shave time and cost from manufacturing processes. “You grow into automation,” Jay Baron, president and C.E.O. of the Center for Automotive Research, told me. “You don’t start with it.” In one of his many public claims about his new plant, Musk expressed confidence that the assembly line would eventually move at about one metre per second, which Baron calculated as one car every fifteen seconds. Such a pace is twenty times the assembly line’s current speed-it’s an almost comical pace, and one that raises serious questions about what Telsa would have to sacrifice to maintain this speed on the line. There are upwards of thirty thousand components in a typical car, and while Teslas have fewer, because its electric-propulsion system is a bit simpler than a gasoline-powered engine, making them is still a complicated process. At the speeds that Musk is anticipating, it would be virtually impossible for Tesla’s network of suppliers to feed the assembly line with parts that meet high-quality standards without interruption. In addition, automakers budget a great deal of time to randomly inspect partially built cars for flaws and errors that could indicate larger problems with other vehicles on the line. Musk hasn’t yet addressed these contingencies, all of which will affect the safety of his vehicles, and Tesla declined to comment further. Although the company’s cars have been given high marks by Consumer Reports and Car and Driver for their on-road performance, Consumer Reports’ 2016 Annual Auto Reliability Survey ranked Tesla twenty-fifth out of twenty-nine brands. The magazine cited a slew of issues with the seventy-five-thousand-dollar (or more) Model X S.U.V. crossover, “including frequent malfunctions of the falcon-wing doors, water leaks, and infotainment and climate-control system problems.” Other concerning information has come out in recent months. As the Wall Street Journal reported, a design engineer for Tesla named Cristina Balan noticed in 2014 that Model S cars on the assembly line had an observable gap between the headliner (the material on the interior of the car’s roof) and the trim on the adjacent support pillar. Worried that other, more critical parts didn’t fit or meet tolerances, and upset that people were paying more than a hundred thousand dollars for shoddy workmanship, Balan told her managers and Musk about the problem. Over the next few months, Balan says she was demoted, harassed, and ultimately forced to quit-even though she had been so integral to the development of Tesla’s power system that her initials were engraved on each battery. Balan is now in arbitration with Tesla over her departure from the company. When Balan first left the company, she didn’t ask for severance pay or even that her sacrificed stock options be allowed to vest, but only for a private meeting with Musk, which was rejected. She told me, “I wanted to alert Elon, without lawyers or V.P.s present, that his company was in trouble. People are afraid of losing their jobs if they speak up and this will get worse as they try to mass-produce automobiles in a company that isn’t ready to do that.” Tesla denies that Balan was forced to quit over the issues she raised. Musk’s optimism about Tesla’s ability to fund both his ambitious factory and the retail network needed to sell his mass-produced automobiles seems to be undermined by the dollars earmarked for the task. The company said in October that it would spend about a billion dollars on capital expenditures, mostly for the new plant, in the fourth quarter of 2016. By contrast, Nissan has invested more than six billion dollars in its Smyrna, Tennessee, factory, which, with an output of more than six hundred thousand cars a year, is the most productive auto plant in the U.S. by volume. Tesla’s planned outlays also appear to give short shrift to the dozens of repair centers that will be required to maintain its customers’ cars; already Tesla owners have complained about delays of a month or more for routine or pressing issues, according to an article in Automotive News.