To hear Tesla’s management talk, you might think that the record-breaking $1-trillion pay package its board just approved for Elon Musk is, in fact, a bargain.

The proposal, which will be presented to Tesla shareholders for their approval on Nov. 6, would require Musk to dramatically remake the electric-car company, turning it into an AI and robotics company that also sells cars.

“Mr. Musk is being asked to transform Tesla and, in so doing, provide transformative technologies to better society as a whole,” the company says in its proxy statement in which the pay deal is disclosed. “No company in the world has similar goals to what is expected under the pay-for-performance structure of the 2025 CEO Performance Award, nor do other companies have compensation plans where pay for their most important executive is entirely contingent on achievement of such daunting and challenging goals.”

Much of the news media has taken as gospel the assertion that the benchmarks Musk must meet by 2035 to collect his fill compensation are indeed “daunting and challenging.” A close read, however, reveals that they’re nothing of the kind.

Most of them are “watered-down versions” of Musk’s own “broken promises,” in the words of Tech Crunch, one of the few tech news sites to have carefully analyzed the pay package.

Tech Crunch is correct. The pay package is replete with smoke and mirrors, its benchmarks painstakingly tweaked to make them much more achievable than they appear on the surface.

Tesla Chair Robyn Denholm, a longtime Musk acolyte, has embarked on a media tour to defend the package ahead of the vote, telling the Financial Times, Wall Street Journal and New York Times that the board judged its terms necessary to motivate Musk to devote his attention to the company over the next decade.

“The board’s opportunity here is to motivate Elon to do impossible things,” Denholm told the Financial Times.

Musk has threatened before to turn his back on Tesla if he isn’t satisfied with his level of authority. Last year, he intimated by tweet that he might not be inclined to develop AI capabilities within Tesla, as opposed to at his other companies, unless the Tesla board granted him a 25% voting control of Tesla.

He’s not there yet — the proxy statement places his share ownership at 19.7% as of Aug. 29. If all the benchmarks in the new package were fully achieved, however, that would bring him shares equivalent to an additional 12% of the company, thereby meeting his demand.

The Tesla board’s faith in Musk is, in a sense, endearing. But it’s also self-defeating. Under his leadership, Tesla’s automotive sales have been slipping, despite price cuts. Competing EV makers have been taking global market share from Tesla, and the trend looks likely to continue. Its highest-profile new product, the widely ridiculed Cybertruck, is a bust, with only 4,306 sold in the second quarter that ended June 30, down 50.8% from the same period a year earlier.

Tesla’s profits per share and revenue have been failing to meet investor expectations. Its share price had lost about 2% this year through Friday, a period in which the Standard & Poor’s 500 index gained about 12%. Tesla shares soared Monday, however, when it was disclosed that Musk had bought $1 billion in Tesla shares after the pay deal was reached. Yet during trading Monday, the shares were still about 12.8% below their peak closing price last year of $479.86, reached on Dec. 17.

A record like this would have a truly independent board hunting for a new chief executive, not crafting a landmark pay package for the old one. But the board could hardly be considered independent.

Among its members are Kimbal Musk, Elon’s brother; James Murdoch (Rupert’s son) and Ira Ehrenpreis, who are personal friends of Musk’s; J.B. Straubel, a former Tesla executive and director of SolarCity, a Musk-controlled solar equipment company that Musk merged into Tesla in 2016; and Denholm, who has said that the money she has earned as a Tesla director has been “life-changing.”

Tesla says the two Musks recused themselves from the board vote on the compensation package, but if you think the result doesn’t reflect exactly what Musk wanted, you should probably think again. Anyway, the other directors approved the deal unanimously.

Now let’s scrutinize that deal. It requires that to collect the full disbursement by 2035 Musk will have to achieve the following: Deliver a total of 20 million vehicles, place 1 million robotaxis into commercial operation, reach 10 million active subscriptions to Tesla’s Full Self Driving (FSD) functionality, deliver 1 million “bots,” increase earnings before taxes and other expenses, or EBITDA, to $400 million a year and increase the company’s market value to $8.5 trillion.

The deal also requires Musk to create a succession plan for CEO, and to “wind down in a timely manner” his political involvement. But it didn’t define “timely.”

On first glance, these do seem to be “daunting and challenging.” However, as tech blogger Will Lockett observes, “each has been twisted and worded in a way in which Musk can achieve them without generating any tangible growth.” I’ve asked Tesla to respond to critiques of the deal terms, including those by Tech Crunch and Lockett, but haven’t received a response.

Let’s start with the 20 million vehicles. The figure includes the roughly 8 million vehicles Tesla has already delivered. Tesla has delivered about 1.8 million vehicles in each of the last two years. At that pace, the company is on track to fill out the rest of the 20-million goal in less than seven years. That’s far short of the goal that Musk had consistently set for the company of selling 20 million vehicles per year by 2030. Rather than a “daunting” challenge, in other words, that one looks like a gimme.

What about placing 1 million robotaxis in operation? Musk has talked up the idea that a robotaxi fleet will be a key to Tesla’s profitability in the future, but this is another walk back of a Musk promise. In 2019, he forecast that 1 million Tesla robotaxis would be on the road within a year. But the company only rolled out 20 or 30 prototype taxis this year, in Austin, Texas, with human drivers behind the wheel.

So that would seem to mean Musk has a long way to go — except that this goal is also not what it seems. The deal defines “robotaxi” as any vehicle that “uses FSD and is used to offer transportation services without a human driver in the vehicle.” The requirement is for a “daily average aggregate number of 1 million vehicles commercially operated by or on behalf of the Company” over any “ three-month period.”

Nothing appears to prevent Tesla from offering paid rides in FSD-equipped cars for a few days here or there over a three-month span. The giveaway, Lockett points out, is pegging this benchmark to the number of vehicles rather than the number of paid drives, a more appropriate metric.

Then there’s the goal of 10 million active subscriptions for FSD. It’s proper to note that around the time it announced the pay deal, Tesla redefined what “FSD” means. Musk long interpreted it as fully autonomous driving, as though the passenger could nap, read a book, play with an iPhone, whatever, leaving the driving to the machine. Now it markets “Full Self Driving (Supervised),” meaning “Your car will be able to drive itself almost anywhere with minimal driver intervention,” obviously not the same thing. The company says that currently enabled FSD features “require active driver supervision and do not make the vehicle autonomous.”

The proxy doesn’t state how Tesla must reach the subscription goal — whether it could, for instance, cut the price of a subscription to pennies, as opposed to the $8,000 it’s charging for the “supervised” FSD. The proxy defines FSD loosely as “an advanced driving system, regardless of the marketing name used, that is capable of performing transportation tasks that provide autonomous or similar functionality under specified driving conditions.”

As for “bots,” 1 million of which will have to be in service, the compensation plan defines them “as any robot or other physical product with mobility using artificial intelligence manufactured by or on behalf of the Company, including Optimus … that substantially performs or provides similar functionality as such robot or other product using artificial intelligence,” not including its cars. Optimus is Tesla’s humanoid robot that has yet to come close to the mobility standards set by robotics firms such as Boston Dynamics.

The proxy is silent on where or how these bots must be marketed. A million Optimus robots could be bought by SpaceX or any other Musk enterprise, placing the goal entirely in his hands. Never mind that humanoid robots have limited utility — the robots that have been successful in industry, including those operating in Musk’s factories, don’t resemble humans like Optimus. They’re custom designed for specific duties, which seldom require arms, legs, or dance moves. And Tesla isn’t at this moment a competitor in that space.

Finally, consider the financial benchmarks — $400 billion in annual pretax earnings and a market capitalization of $8.5 trillion. The deal doesn’t adjust either figure for inflation, so that could do part of the job of increasing Tesla’s adjusted annual earnings from the $16.6 billion it recorded last year. But there are a few ways Musk could goose Tesla’s earnings in the next 10 years, including through the acquisition of his artificial intelligence company, xAI, or SpaceX.

It’s important to note that Musk doesn’t have to reach all these milestones to collect multiple billions of dollars in pay over the coming decade. The deal is divided into 12 equal “tranches” of restricted shares, and he can start collecting as soon as he reaches a minimum adjusted income, starting with $50 billion, and pairs that with an unmet operational goal such as car sales and FSD subscriptions.

The question is whether reaching any of these goals actually puts money in Tesla shareholders’ pockets via tangible growth. If the Tesla board really wanted to hold Musk’s feet to the fire, it would have produced a very different package.