Musk is a complete genius. Just look at how well Tesla has done over the last few years. Sales are way down, and operational profits are now virtually non-existent. The company’s stock price has fluctuated but has been overall completely stagnant for five years. The models that were expected to dominate the EV market have either flopped or failed to materialise. Tesla no longer has a lead in the EV market, and competitors are actively leapfrogging it, given that Tesla’s technology has failed to develop for years now. But worst of all, Tesla is still light years away from the promises of fully autonomous cars and Robotaxis, which Musk made a decade ago. Oh, and now the entire world associates Tesla with Nazis, which has completely destroyed tens, or possibly hundreds, of billions of dollars in brand value and driven huge swathes of customers away. So, what does Tesla do to punish this “genius”? Offer him $1 trillion…
Tesla’s board recently offered Musk a new pay packet, and it is somehow even more insane, pathetic and stupid than his old $56 billion one.
In order to receive the pay packet, Musk has to meet specific targets over the next ten years (up until 2035). They are listed below:
Deliver 20 million total vehicles.
Enter one million Robotaxis into commercial operation.
Reach ten million active subscriptions to Tesla’s Full Self-Driving (FSD).
Deliver one million Bots (Tesla defines “Bot” as “any robot or other physical product with mobility using artificial intelligence manufactured by or on behalf of the company” — yet, somehow, the company’s vehicles do not count).
Increase revenue from the current $17 billion to $400 billion.
Create a succession plan for the CEO position.
Provide Tesla with some assurance that “Musk’s involvement with the political sphere [will] wind down in a timely manner.”
Ensure Tesla reaches an $8.5 trillion valuation.
If Musk can achieve all of these targets, Tesla will give him more ownership, increasing his stake to 27%, which at an $8.5 trillion valuation is equivalent to a $1 trillion payday. The idea of this ludicrous payout is to motivate Musk to focus on Tesla — though I don’t think they realise quite how lazy that makes him sound.
Now, you might be thinking, “Well, fair enough, these targets seem incredibly difficult. If he can reach them, why not pay him well?” But here is the thing: each one of those targets is dramatically watered down from Musk’s own promises, and they are all worded in such a way to make achieving them far easier than you might think.
So, let’s break down each one of these targets.
20 Million Total Car Sales
This target is possibly the most frustratingly easy for Musk to reach and demonstrates just how much the company has given up on EVs.
Not very long ago, Musk promised that Tesla would sell 20 million vehicles per year and guaranteed vehicle sales growth of 50% per year over the next decade. However, thanks to the 4680 becoming obsolete before it has officially hit the market, the competition taking huge leaps beyond Tesla, the Model 2’s cancellation or severe delay, Tesla’s sales slumping even before Musk entered politics, and Tesla’s valuation becoming increasingly dependent on AI speculation, this path was abandoned in favour of developing FSD, Robotaxis and Optimus bots.
Tesla has already sold a total of eight million vehicles, and despite sales plummeting, it continues to chug along at two million sales per year. So if nothing changes, Tesla will hit 20 million total vehicles sold by 2031. That is four years before the deadline.
The only thing Tesla is asking of Musk is that he doesn’t tank vehicle sales much further or completely abandon solid business fundamentals in favour of wildly speculative AI valuations.
Not only is this a dramatic watering down of Musk’s own promises, but this target is actually below the sales numbers Tesla is already achieving. This isn’t a goal; it is asking the bare minimum.
One Million Active Commercial Robotaxis
On its face, this target seems far harder to reach. After all, as I have written about time and time again, Tesla’s FSD system is wildly unreliable and totally unsuitable for this kind of operation. It has no system redundancy, which means its AI must be completely flawless in order to function, and, unsurprisingly, this is impossible to build (read more here and here). This is why open third-party data suggests FSD is more than 99.87% less safe than Musk promised (read more here), even when the safety levels Musk promised were already unfit for Robotaxi operations. That is why Tesla recently moved the safety driver for its fledgling Robotaxi service (which consists of only around 20 vehicles) from the passenger seat to the driver’s seat, given that they are expected to take control so often.
So, Musk needs to make FSD more than 100 times safer and scale Tesla’s Robotaxi fleet by a factor of 50,000 to even hit this mark. That will require a complete redesign of FSD and a rapid rollout, which will cost tens of billions of dollars and is nearly impossible. Right?
Well, no. The fine print of the pay packet says the target is to reach a “daily average aggregate” of one million Robotaxis “commercially operated by or on behalf of [Tesla] over a consecutive three-month period, as part of a transportation service.” And it defines “Robotaxi” as any Tesla vehicle using FSD to offer rides to people.
Musk has also said that Cybercab production will begin in 2026 and rapidly scale to two million units per year, with Tesla holding onto the majority of them. So, reaching one million vehicles capable of operating as a Robotaxi over a three-month period is actually far below Musk’s stated goal.
And it gets worse, because the fine print doesn’t describe what “commercially operated” means. Is that a car being used for the entire time? Or does it also count when it is on standby? So, these Robotaxis don’t have to be used every single day for that three-month period. Furthermore, it doesn’t specify that FSD has to be unsupervised. In other words, an Uber driver using FSD but still needing to regularly take over could count towards this one million number. Or, if demand for that is low, Tesla could easily offer their own low-use system and just cycle through one million vehicles, with each vehicle only doing a few trips over a three-month period.
What is telling is that the number of Robotaxis is the element being prioritised, not how many trips are offered. The money and value in Robotaxis are in the number of trips they can deliver, as it is much higher than a human-driven taxi. That is why Waymo is such a powerhouse with only a few thousand vehicles.
If Tesla was serious about growing its Robotaxi business, its main priority would have been the number of trips offered per week or the total revenue from rides, not the number of operational vehicles, because that number is irrelevant, encourages inefficiency, and is far easier to manipulate.
And, let’s be honest, Musk has a rich history of this kind of manipulation.
Ten Million Active FSD Subscriptions
Okay, well, this target only proves my point, right? If Tesla wants to grow its FSD business, it isn’t about vehicles sold but the number of subscriptions. So, surely this is a much more concrete and difficult goal to hit?
Ha! Not at all.
Yes, FSD is a sales flop. Third-party data suggests that only 2% of people offered a free trial actually pulled the trigger and paid the massively discounted $100-per-month subscription, which isn’t surprising, considering how badly it drives. In fact, major industry players are currently pulling their autonomous systems from the market, as there is simply not enough demand for expensive systems that only minorly increase customer convenience. So, this is an industry-wide problem, and increasing FSD’s subscriptions to ten million seems frankly impossible as a result. Musk will have to make it fully autonomous and completely unsupervised, as well as convince most Tesla drivers to pay for that service.
But there is a caveat.
Tesla recently changed what FSD means. Since 2016, if you bought FSD or paid for the subscription, Tesla promised that you would eventually get a fully autonomous car. That is despite the fact that Tesla has already acknowledged that vehicles produced between 2016 and 2023 don’t have sufficient hardware to deliver unsupervised self-driving. Or the fact that all the evidence points to the AI being so fundamentally inaccurate that it can’t make the current FSD system safe enough for this kind of use.
So, as Electrek reported, Tesla recently changed the terms and conditions of what customers buy when they sign up for FSD. Rather than promising to deliver a fully autonomous vehicle, Tesla is instead just selling it as a driver assistance system.
(As a side note, I highly suspect this is because Tesla has realised FSD isn’t suitable for Robotaxi operations and is working on a much more substantial system to ensure the system is safe enough to compete with the likes of Waymo. Unfortunately, the caveat is that this system won’t be backwards compatible with older Teslas, as it will require far more sensors and far better computer systems.)
Suddenly, having ten million active FSD subscriptions means something totally different. It doesn’t mean that ten million people will pay for a fully autonomous car at all. There are also no dependent conditions, like how much an FSD subscription is sold for, or if packaging it up with other services counts.
That makes it really easy to meet this target. Tesla has a direct-to-consumer model, so it can use mandatory subscriptions to keep its vehicles operational or paywall key features behind a subscription. What if Tesla bundles FSD with other features like heated seats, the EV charger navigation, faster charging speeds, their in-house insurance or a servicing plan, and sells it for a few measly dollars per month to buyers? With two million Teslas being delivered each year and eight million already on the road, they could easily hit their ten million mark way before 2035.
So, by changing what FSD means and deliberately leaving the conditions open to loopholes, this is another target that is piss-easy for Musk to hit.
One Million Bots Delivered
Back in 2023, FSD was expected to drive Tesla’s growth. Cathy Woods even suggested that by 2027, it would increase Tesla’s revenue to $400 billion, or 23 times its current revenue, and its share price to $2,000, or six times the current price. Musk publicly agreed with these demonstrably unhinged claims. But the past few years have removed the sheen from FSD, mostly due to the system’s monumental flaws and Tesla’s stupid business plan becoming more apparent. So now, it’s Musk’s humanoid robot Optimus that will drive growth, with Musk himself saying they will soon generate 80% of Tesla’s revenue and increase Tesla’s valuation to $5 trillion, or five times its current value.
But, just a year ago, Musk claimed Tesla would sell one million Optimus robots per year by 2029. He even said Tesla could sell 20 billion of them in total. So again, this target is heavily watered down from Musk’s own promises.
However, there is a significant problem here: humanoid robots make no sense. They are hard to program, inefficient, unreliable, require high levels of supervision, and are expensive. And that is true for significantly more advanced humanoid robots like those from Boston Dynamics (read more here). By comparison, even after years of development, Optimus is no more capable than Honda’s 20-year-old ASIMO robot. The argument that Tesla’s AI will give them an advantage here is also laughable. A humanoid robot has many more variables and will be asked to do a much broader range of tasks than their, by comparison, highly constrained application of FSD, which Tesla finds difficult enough to make functional (even before you consider the fact that they are already bumping into the limits of AI technology).
Quite simply, Optimus physically can’t achieve what Musk is promising. But to Tesla, that doesn’t matter.
Again, there are no conditions here — the only thing the board is asking Musk to do is deliver a million bots. But, like the Robotaxi, the true metric for growth is how many are being actively used and for how long. There are also no conditions on how these deliveries happen.
For example, Musk’s own companies could order them. While they are painfully ineffective at factory work, Tesla and SpaceX combined have nearly 150,000 factory workers. Optimus can’t autonomously do these jobs, and using it in these applications will make these factory lines less efficient and won’t actually reduce the workforce, as these same people will need to supervise Optimus. But that doesn’t matter; they don’t have to use them, just order them and test them. Then, when a new version of Optimus comes along, throw the old ones away and try the new ones, and so on. Musk could easily shift several hundred thousand of these bots like this.
As for the rest, Musk is preparing to construct a variety of data centres for xAI, Tesla and SpaceX in the coming years. There are specialised robots that can build, service and maintain these sprawling sites far more efficiently than a general-purpose Optimus. But again, he can just stuff these data centres full of Optimus bots to get the numbers up, and oh, would you look at that? Over ten years, Musk has delivered a million humanoid robots without selling a single one to the public.
Again, if Tesla was serious about Optimus, this target would prioritise revenue generated by non-Tesla Optimus sales, like the number of daily hours these robots worked in key jobs, or something similar. Basing it on deliveries far below what Musk promised is just pathetic.
Revenue Increase to $400 Billion Annually
Keep reading with a 7-day free trial
Subscribe to Will Lockett's Newsletter to keep reading this post and get 7 days of free access to the full post archives.