A16Z: The sun bears witness, SpaceX is worth 7.5 trillion

Original Author: @mikemcg0 and @pmarca

Original Compilation: Yu Eleven

Musk's compensation plan at SpaceX revolves around two goals.

The first goal is: If the company's valuation reaches $7.5 trillion and a permanent human colony of at least 1 million people is established on Mars, he will receive the first reward.

The second goal is: If SpaceX operates data centers in space that consume at least 100 terawatts of power—this number exceeds the total power consumption of all data centers on Earth by 1000 times—he will receive the second reward.

If neither goal is achieved, Musk will receive nothing, except for the $54,080 annual salary he has been receiving since 2019.

The board members who signed this compensation plan have witnessed one thing over the past twenty years:

Musk has repeatedly made seemingly impossible predictions about SpaceX, and those predictions have repeatedly become reality.

He said that SpaceX would send humans into orbit—before that, no private company had done so. Now, SpaceX regularly transports astronauts for NASA.

He said that SpaceX would land and reuse orbital-class rockets—before that, the entire industry viewed boosters as disposable. Now, SpaceX has completed hundreds of recoveries and reuses.

He said that a satellite internet business could be worth tens of billions of dollars—before that, satellite internet was almost a graveyard for bankrupt companies. Now, Starlink's revenue has grown from zero to $11.4 billion in just a few years.

These predictions are often aggressive in their timelines, but they have almost never missed the mark in terms of direction.

And SpaceX's original mission, written in 2002, is to make humanity a multiplanetary species.

Thus, the board tied his compensation to this mission itself.

If this mission sounds like science fiction, it may be because it indeed comes from science fiction.

---

Iain M. Banks and the Blueprint of "The Culture"

Iain M. Banks spent twenty-five years writing a world called "The Culture."

By most reasonable standards, it may be the best utopian society ever imagined by humanity.

There, humans live alongside superintelligent AIs known as Minds. The Minds are responsible for running enormous orbital habitats. The relationship between humans and AIs is not one of enslavement or competition, but partnership.

No one is forced to work.
No one goes hungry.
The Minds handle the astonishing computational load required to run space cities.

And humans are responsible for continuing to be human.

This task itself is a full-time job.

The three autonomous drones SpaceX uses to recover Falcon 9 boosters at sea are named after conscious starships from Banks' novels:

• Of Course I Still Love You
• Just Read the Instructions
• A Shortfall of Gravitas

In an interview at the 2023 UK AI Safety Summit, Musk was asked: What should a good AI future look like?

He replied:

"Banks' Culture series is the best imagination of the AI future so far. Nothing else comes close to helping you understand what a fairly utopian, or progressive utopian AI future could be."

He has been telling us what he wants to build by naming those landing platforms.

Caption: "Of Course I Still Love You" caught the Falcon 9 first stage booster on April 8, 2016. This was the first successful drone ship landing in history and the moment reusable orbital spaceflight transitioned from theory to reality. The ship's name comes from a conscious starship in Iain M. Banks' Culture series. (Image: SpaceX)

But "The Culture" is not a frictionless paradise.

Banks' novels are filled with war, intrigue, and moral complexity. It is a utopia because civilization has solved the prerequisites for survival well enough that trillions of humans can finally attend to what Banks calls "the truly important things in life":

Exercise, games, love, studying dead languages, barbaric societies, impossible problems, and climbing mountains without a safety net.

Such a future has four prerequisites.

First, the ability to obtain a meaningful portion of a star's energy output—this is several orders of magnitude higher than the energy produced by human civilization today.

Second, large-scale physical intelligence: machines capable of building, mining, refining, and repairing anything without human intervention, and able to do so anywhere.

Third, cheap digital intelligence that exceeds biological intelligence.

Fourth, the ability to cheaply, frequently, and reliably send mass off Earth. Because all of the above cannot scale relying solely on Earth itself.

---

Backtracking from the Future

Most analyses of SpaceX are forward-looking:

Rockets, satellites, contracts, revenue.

But to see what is really happening, a more useful approach is to start from the end and work backward.

---

Martian City

The operational goal is:

To establish a 1 million population self-sustaining city on Mars within the lifetime of people alive today.

The real difficulty lies in "self-sustaining."

This means: If Earth stops sending supplies, the city must still survive.

It must produce everything itself:

Food, water, air, energy, medicine, machinery, and eventually more humans.

According to SpaceX's own estimates, sending 1 million people and millions of tons of cargo to Mars within decades will require thousands of Starship flights; during each transfer window, more than ten launches must occur each day.

These windows are determined by Earth-Mars orbital mechanics, are only a few weeks wide, and open only once every 26 months.

Caption: SpaceX's rendering of the Martian city. (Image: SpaceX)

---

Lunar City

The lunar city is a closer, easier rehearsal.

There is ice in the permanently shadowed craters at the lunar south pole, and certain ridges receive continuous sunlight, making it a natural location for a base.

But what Musk is talking about is not just a research outpost.

He envisions building factories on the Moon to produce AI satellites and launching them one by one into space using mass drivers.

The mass driver is also a concept Musk borrowed from science fiction. It is an electromagnetic launch system that uses the Moon's one-sixth gravity and lack of atmosphere to throw solar satellites into deep space at an industrial scale.

These satellites can be manufactured on the Moon because lunar regolith contains about 20% silicon and 10% aluminum by weight—these are the two main inputs for solar cells and satellite structures.

Musk explains: "If you want to exceed a scale of 1 terawatt per year, you have to go to the Moon."

Caption: SpaceX's rendering of the mass driver at Moonbase Alpha, used to launch AI satellites, i.e., data centers, manufactured on the Moon into orbit. (Image: SpaceX)

---

Orbital Data Centers

Musk bets that:

In a few years, from an economic standpoint, the most suitable place to place AI data centers will be in space.

The bottleneck for AI is energy. Except for China, energy supply growth is very limited, while the demand for AI computing power is growing exponentially.

Solar panels in orbit provide 4 to 10 times the power of solar panels on the ground. The specific multiple depends on how sunny the ground location is.

The reason is simple:

There is no atmosphere in space, no day-night cycle, no clouds, and no seasons.

NASA figured this out decades ago. Now, rockets are finally cheap enough to make it a reality.

Musk predicts that in five years, the AI computing power launched into orbit by SpaceX each year will exceed the total installed computing power on Earth.

This is why SpaceX merged with xAI in February.

Rockets and intelligence are becoming the same issue.

---

Starship: The Vehicle for Everything Upstream

Starship is the vehicle that makes everything upstream possible.

The Starship V3, which had its maiden flight this year, is the largest and most powerful rocket ever built by humanity. It is taller than a 40-story building and has more than twice the thrust of the Saturn V rocket that sent astronauts to the Moon.

According to NASA statistics, the historical cost of reaching orbit is about $18,500 per kilogram.

In 2010, the first Falcon 9 reduced this cost by about 85%, to around $2,700 per kilogram.

In 2018, Falcon Heavy further reduced it to about $1,400 per kilogram.

And Starship, as the world's first fully and rapidly reusable spacecraft, aims to bring the cost down further to $100 to $500 per kilogram.

What used to cost billions for a single launch is now becoming a business in the tens of millions of dollars.

---

Starlink: The Cash Wheel

Starlink is the cash wheel that helps pay for everything else.

According to SpaceX's IPO documents, the connectivity business segment—almost entirely Starlink—is expected to bring in $11.4 billion in revenue in 2025, a year-on-year growth of about 50%, with an adjusted EBITDA margin exceeding 60%.

As of March 2026, Starlink has 10.3 million users in 164 countries, operating on over 9,600 satellites.

Starlink was initially just a side project to fill the company's own launch capacity; now, it is becoming one of the great consumer businesses in history.

In 2019, when a16z conducted due diligence on SpaceX, several people told us that this economic model would never work.

The reason was that Starlink's terminal antennas required technology previously used only in F-22 fighter jets and naval destroyers, and this technology had never been mass-produced for consumers.

SpaceX's first batch of terminals had a manufacturing cost of about $3,000, but they sold for only $499.

But they eventually brought the manufacturing cost down and proved the skeptics wrong.

---

Falcon 9: The Workhorse Buying Time for the Future

Falcon 9 is the workhorse buying time for everything else.

It is the only orbital-class booster on Earth that is reused at scale. A single booster can typically fly more than twenty times before retirement.

In 2025, SpaceX launched 83% of the total mass into orbit globally.

Despite everyone else having a half-century head start, SpaceX's payloads launched into orbit now exceed the total of all other forces in the world.

This is the stack from top to bottom.

Generations later, a Culture-like future resides at the top.

Falcon 9 and Starlink sit at the bottom, paying today's bills.

Each layer makes the next layer possible.

SpaceX CFO Bret Johnsen described the internal feeling at the company:

"[Musk] created a culture: you first set an initial goal that seems incredibly bold, and then step by step, you find yourself moving toward something that is absolutely achievable...

Take going to Mars as an example. When I first arrived in 2011, whenever people talked about Mars and multiplanetary species, others would roll their eyes. Now when we talk about it, the reaction has changed to: 'What year?'

I think Elon’s greatest strength is that he sets these goals and builds a really great business model around every piece of intellectual property needed to achieve the ultimate goal."

---

The Idiot Index and the Algorithm

Musk initially did not intend to start a rocket company.

In 2001, at 30 years old, Musk was contemplating what to do after selling PayPal.

He had always been interested in space. When he looked for NASA's plans to send humans to Mars, he was surprised to find that there was no such plan.

So he designed a proposal:

To send a small greenhouse to Mars and transmit photos back to Earth.

His idea was that if people saw a green sprout appearing on the dead red planet, it might reignite public interest in space and stimulate political will to fund a real Mars program.

He just needed a rocket to send the greenhouse.

Later that year, he went to Moscow to buy a refurbished intercontinental ballistic missile. This was the first of two trips to Moscow.

It is said that those meetings were filled with vodka and posturing.

Musk's best friend from the University of Pennsylvania, Adeo Ressi, went along. He told Esquire in 2012:

"We would walk into a small room, and everyone had their own bottle of vodka in front of them."

The Russians did not take Musk seriously.

At one point, a chief designer even spat at Musk and his team in contempt.

The second trip to Moscow was in February. Musk asked how much a missile would cost.

The answer was: $8 million each.

Musk countered: Two for $8 million.

Musk's space advisor Jim Cantrell remembers that the other side said something like:

"Young man, that's not going to happen."

And implied that he had no money at all.

Musk judged that they were not serious, so he got up to leave.

Cantrell thought that the trip was over.

On the return flight, he and Mike Griffin, who later became NASA's administrator and was also traveling as an advisor, ordered drinks and toasted to finally leaving Moscow.

Musk sat in the row in front of them, hunched over looking at his laptop.

Then he turned around:

"Hey, guys, I think we can build this rocket ourselves."

He showed them a spreadsheet listing the raw materials needed for the rocket: aluminum, titanium, copper, carbon fiber, and the cost of each material.

The cost of these materials accounted for only 2% of the quote.

As Musk later said:

"It was clear that you just needed to figure out a smart way to combine these materials into the shape of a rocket."

Within months, Musk decided to risk $100 million to start a rocket company. This was more than half of the approximately $180 million he received from selling PayPal.

SpaceX was thus founded in a warehouse in El Segundo, California.

He invited five people to form the founding team.

Three declined, including Cantrell and Griffin.

The two who agreed were:

• Tom Mueller, who later became Vice President of Propulsion and the company's first employee;
• Chris Thompson, the second employee, responsible for operations and production.

Musk later joked:

"SpaceX in 2002 basically had a carpet and a Mexican mariachi band. That was it. You can see, I’m a dancing machine."

Years later, Musk referred to the diagnostic tool behind that spreadsheet as the "idiot index."

If the cost of a part is high relative to its raw material cost, then either you are an idiot, or you are working with an idiot.

This sounds like a joke, but it is the foundation of SpaceX's strategy.

Every part SpaceX procures comes with an "idiot index" calculation.

One of the most legendary stories from the early days of the company involves Steve Davis.

Davis joined SpaceX directly after graduating from Stanford and was the company's 14th employee. His task was to procure an actuator to control the steering of the upper stage of the Falcon 1 rocket.

When he reported that traditional aerospace suppliers were charging $120,000, Musk laughed.

Musk told him that the complexity of this component would not exceed that of a garage door opener and gave him a budget of $5,000 to build it from scratch.

Biographer Ashlee Vance recorded that Davis spent nine months iterating on the design and ultimately created a usable actuator that cost only $3,900.

When Davis sent Musk the breakdown of this victory's technology, Musk replied with just two letters:

"Ok."

To push the idiot index toward its theoretical limit, you must vertically integrate and control the process end-to-end.

But vertical integration incurs fixed costs that are only economical at high volumes.

And to achieve high volume in the rocket business, you must break the traditional operating model of the industry.

Traditional launch service providers, like ULA and Arianespace, treat each mission as a custom project.

Customers specify the orbit, payload, and integration requirements, and the launch service provider designs a custom mission around the satellite.

This model assumes:

There are only a few launches per year, and each mission is extremely costly.

It makes scalable manufacturing impossible.

SpaceX does the opposite.

They published a Falcon user guide that defines the precise specifications of the rocket and tells customers:

Please design your satellites to fit our rocket.

At the time, this was considered very radical and caused SpaceX to lose some early business.

But it opened the manufacturing flywheel.

Standardization and reusability reinforce each other.

Because every Falcon 9 is the same, recovered boosters can be turned back into a complete, qualified product ready to fly again.

The first Falcon 9 booster to fly twice was successfully reused in 2017.

By 2020, a single booster could fly five times.

By 2021, it could fly ten times.

Today, the record holder has flown 35 times.

This reusability has changed the economics of spaceflight, and it is hard to see how competitors can catch up.

In 2021, Musk estimated that the marginal launch cost of Falcon 9 to put 15 tons into orbit (excluding indirect cost allocation) was about $15 million. He said this was about one-half to one-third of the cost of alternatives.

Today, SpaceX launches a rocket with reusable boosters every two to three days, while competitors can only launch a few custom rockets a year.

But SpaceX's advantage is not just economies of scale, vertical integration, and better strategy.

It also includes speed and culture.

Traditional aerospace companies eliminate uncertainty through analysis.

NASA once described Boeing's commercial crew program in polite terms:

"Employing a mature systems engineering approach, with upfront investment in engineering research and analysis to mature system design before manufacturing and testing."

Measure twice, cut once.

SpaceX reversed this order.

The company builds many cheap prototypes, pushes them to failure, learns from the failures, and iterates quickly.

The testing program for Starship has produced spectacular explosions that may surpass any rocket project in history.

But each failure is a data point of where reality diverges from the model.

This contrast is very clear to those who have worked in both worlds.

Garrett Reisman is a NASA astronaut who flew on two shuttle missions. In 2011, he left NASA to join SpaceX as a senior engineer.

He described the mainstream view of SpaceX at NASA at the time:

"They are cowboys; they are dangerous; they will kill people."

But what really changed his view was seeing how SpaceX operates.

"The things they could produce in a month, NASA might take a year. We were shocked."

The clearest example is the Falcon 1 project.

Between 2006 and 2008, SpaceX launched four Falcon 1 rockets from a small atoll in the Pacific called Kwajalein.

The first three failed.

But each failure was different and provided learning opportunities:

• The first, fuel leak;
• The second, abnormal propellant oscillation;
• The third, residual engine thrust caused a collision during separation.

By September 2008, the company had enough money left for one more launch.

And this was not the only company Musk had standing on the edge of a cliff.

His electric vehicle company Tesla was also just weeks away from bankruptcy.

He had to decide: Should he concentrate the remaining PayPal cash on one company or split it between the two?

Musk recalls:

"That was a very difficult decision. In the end, I decided to split my remaining money and try to keep both companies alive. But that could have been a very bad decision that led to both companies dying together.

I never thought I would have a mental breakdown, but I was really close."

He couldn't choose because, in his worldview, both missions were indispensable:

Tesla aims to accelerate the world's transition to sustainable energy.
SpaceX aims to make humanity a multiplanetary species.

Musk's then-fiancée Talulah Riley said in the BBC documentary "The Elon Musk Show": "All available resources had to be put into the companies. He gave me the chance to walk away. He said: 'The next part will be the hardest, you don’t have to stay and go through it with me.'"

Caption: In 2006, Elon Musk inspects the wreckage of the first Falcon 1 on Omelek Island. (Image: Hans Koenigsmann)

The fourth launch succeeded.

That December, just weeks before SpaceX was about to run out of cash, NASA awarded the company a $1.6 billion cargo contract.

When NASA called to inform Musk, he was hit by a huge emotional release and blurted out:

"I love you guys."

The pattern formed from rapid failure and rapid correction later became the culture of every SpaceX project.

This is also why today SpaceX can iterate quickly between two Starship test flights, while traditional aerospace projects often take years from one flight anomaly to redesigning the vehicle.

The reason this approach is more effective than alternatives is:

For problems you do not fully understand yet, you cannot arrive at a perfect solution just by thinking.

Reality is the only sufficiently qualified validator.

The key is to lower the cost of consulting reality to a level that allows frequent consultation.

---

SpaceX's "Algorithm"

The above is the iterative loop of SpaceX told through stories.

But it also has a written version.

Over the past twenty years, Musk has encoded SpaceX's methods into a five-step operational process, internally referred to as "The Algorithm."

Tim Berry, who worked at SpaceX for ten years and led the production team for Falcon 9 and Falcon Heavy upper stages, said this method has "drilled into our brains."

Walter Isaacson published its standard version in Musk's biography:

1. Question every requirement

Every requirement should come with the name of the person who made the request.

You should not accept statements like "this requirement comes from the legal department" or "this requirement comes from the safety department."

You need to know who actually made the request, and question it regardless of how smart that person is.

Requests from smart people are the most dangerous because people are least willing to question them.

Then, make those requirements less stupid.

2. Remove everything that can be removed

You may have to add them back later.

In fact, if you do not end up adding back at least 10% of what was removed, it means you haven't removed enough.

3. Simplify and optimize

This step should come after the second step.

A common mistake is to simplify and optimize a part or process that should not exist in the first place.

4. Accelerate the loop speed

Every process can be accelerated.

But this should only be done after completing the first three steps.

Musk said he made a mistake in the Tesla factory: spending a lot of time accelerating certain processes, only to realize later that those processes should have been removed.

5. Automate

Automation comes last.

Tesla has made mistakes in Nevada and Fremont by trying to automate from the start instead of first questioning requirements, removing parts and processes, and shaking out bugs.

Most engineering organizations jump straight to the fifth step.

They try to automate a process that should not exist.

SpaceX runs these steps in order every time, in every part of the company.

When "The Algorithm" runs on a piece of hardware enough times, it starts to look unlike anything in the industry.

Caption: SpaceX Raptor engine generations, from V1 to V3. (Image: SpaceX)

Raptor 3 is the product of a team iterating around the same engine for ten years.

It produces 22% more thrust than Raptor 2, weighs 40% less, and does not require a heat shield.

The reason is that the plumbing and wiring that used to hang outside the engine have been integrated into the metal structure of the engine through 3D printing.

Musk said:

"Simplifying the Raptor engine, internalizing the secondary flow paths, and adding the workload needed for regenerative cooling of exposed components is astounding. It is close to the limits of known physics."

No known engine project in aerospace history has iterated this quickly.

The space shuttle's main engine has essentially flown the same design for the last thirty years.

The RD-180 that powers the Atlas V is a derivative of an engine designed in the 1970s.

And SpaceX has done three complete redesigns of Raptor in less than a decade, each version significantly better than the last.

The same philosophy applies to people.

By mid-2018, Falcon 9 reusability had entered a reliable rhythm, and Musk turned his attention to the satellite internet constellation that would ultimately fund all upstream work.

The Starlink team is located in Redmond, Washington, and many senior engineers come from Microsoft, with a development pace slower than Musk hoped.

In June, he flew to Redmond and fired the senior leadership team.

He then transferred young star engineers from the rocket department and gave them a year to launch the first batch of operational satellites.

This is a brutal management style. From media reports about the firings, the department seemed to be collapsing.

But 11 months later, in May 2019, the first batch of Starlink satellites was launched.

Musk cleared the bottleneck and then turned to the next problem.

He manages everything this way.

In 2018, Tesla was in the "production hell" of Model 3 ramp-up, and the cash burn rate threatened survival. Musk literally moved into the factory.

Years later, he recalled:

"I lived in the Fremont factory and the Nevada factory for three years. I slept on the floor under my desk so that the whole team could see me during shift changes.

This was important because if the team thinks their leader is off enjoying themselves on a tropical island drinking Mai Tais, they will be demoralized.

Because they could see me sleeping on the floor during shift changes, they knew I was there. This made a huge difference, and they gave it their all."

Later, he turned this into a company-wide rule:

The higher the position, the more visible the presence must be.

To find someone comparable to Musk's CEO operating style, one must return to the era of industrialists in the late 19th and early 20th centuries:

Henry Ford, Andrew Carnegie, Thomas Watson, Andrew Mellon, Cornelius Vanderbilt.

What is unique about Musk's operating style is his relationship with work.

It is said that he shows up at each of his companies every week, identifies the biggest problem, and then fixes it.

52 weeks a year, he does this every week.

So theoretically, each company solves 52 of its biggest problems in that year.

An engineer who joined SpaceX from another aerospace company described this experience as:

"It felt like being thrown into a shocking capability zone. Everyone around you is absolutely competent."

---

Constellation

SpaceX looks like a company.

But a more useful way to view it is as a central node in a constellation of companies.

These companies are run by the same person, built toward the same long-term mission, and are almost impossible to separate from each other.

Over the past twenty years, Musk has assembled a group of companies. Each one is solving a constraint that would otherwise become a bottleneck for the others.

Now, they are starting to compound on each other.

SpaceX's merger with xAI in February is a microcosm of what SpaceX is becoming.

If computing power ultimately goes into orbit—this is Musk's bet—then SpaceX has the most credible path to deploy it at the scale required by AI.

Sending mass into orbit and mass-producing intelligence may be the two most critical capabilities for the coming decades.

Now, they are reinforcing each other under the same roof.

xAI brought Grok, a cutting-edge model, and has a unique position in real-time information by accessing X's real-time data stream.

It also brought engineers to build the Colossus 1 and Colossus 2 supercomputers. The speed of these engineers exceeds the imagination of many in the industry.

Caption: Colossus 1. (Image: xAI)

The construction of Colossus deserves a closer look.

xAI took over an old factory in Memphis and got 100,000 GPUs training in just 122 days.

Once the racks started arriving, the entire cluster was running in just 19 days.

NVIDIA CEO Jensen Huang commented on Musk:

"From concept to building a massive, liquid-cooled, powered, permitted factory, and completing it in that timeframe, is superhuman.

To my knowledge, there is only one person who can do this.

What they have accomplished is unique. No one has ever done this. 100,000 GPUs, as a cluster, will easily be the fastest supercomputer on Earth in 2024.

This usually requires three years of planning, then equipment delivery, and another year to get everything running."

A project that would take at least four years for others in the industry, Musk and the xAI team completed in four months.

In May of this year, Anthropic agreed to pay SpaceX $1.25 billion per month for access to all of Colossus 1's computing power.

A few weeks later, in a revision of the IPO documents, SpaceX disclosed that Google would pay $920 million per month for access to 110,000 GPUs, about half of the computing power obtained by Anthropic.

These two deals represent approximately $26 billion in annual revenue.

And this is just two customers paying for a business that did not exist before SpaceX absorbed xAI earlier this year.

Chips, power, and land are scarce.

SpaceX is becoming one of the few companies with enough AI infrastructure to both rent out computing power and pursue its ambition to build leading-edge models.

What xAI gains from SpaceX is a more durable solution to the power constraints that Musk believes will limit AI in the coming years.

To produce enough power to meet the anticipated demand for intelligence requires grid expansion, new power plants, and years of permitting processes, and the industry does not have that much time.

In his view, orbital solar power is the way out because it is virtually limitless.

And SpaceX is the only company with the means to send computing power up at scale.

Whether he is correct is one of the most important open questions in the tech field.

But SpaceX's IPO documents show that the company takes this bet very seriously: it expects AI to be the largest market for the company in the future, far exceeding other markets.

The space business that built the company seems almost like a rounding error compared to these ambitions.

---

Tesla: Another Core Piece of the Constellation

Tesla is another important component of this constellation.

Its integration with SpaceX happens in another way.

Tesla and SpaceX share founders, talent pools, operational cultures, and increasingly overlapping technology roadmaps.

Tesla provides three things to the SpaceX-xAI side of the constellation.

First, chips.

AI5, AI6, and Dojo3 are all designed internally at Tesla.

Musk has made it clear that these chips are not just for cars, but are components of a broader constellation computing stack.

AI5 handles autonomous driving inference.
AI6 is designed for Optimus and AI data centers.
Dojo3 is engineered for orbital computing power, in conjunction with the planned AI7.

Second, robots.

Tesla's bet is that Optimus will become the physical AI layer for factories, warehouses, and homes, handling environments that hope to operate without human labor, ultimately serving the lunar and Martian cities Musk envisions.

Third, solar energy.

Musk said that Tesla and SpaceX are each building toward an annual 100 gigawatts of solar cell production capacity to support AI construction on Earth and in orbit.

Then there is TeraFab.

In April of this year, Tesla revealed that it had begun ordering equipment for a research semiconductor wafer fab within its Giga Texas campus.

Musk told investors on Tesla's Q1 2026 call:

"We expect this to be a roughly $3 billion project, with the potential for several thousand wafers per month."

SpaceX is separately funding the construction of a much larger facility, designed to reach a capacity of about 1 million wafers per month once mature.

The reason is that no existing fab can scale at the speed Musk envisions.

And the scale he envisions is measured in gigawatts.

Last week, Musk said: "This is not something we are promising to do. This is something we will try to do and believe we can probably do: by the end of next year, achieving an annualized rate of about 1 gigawatt in space AI computing power.

Then, from a vision standpoint, scaling up an order of magnitude each year.

That is, in two and a half years, achieving an annualized 10 gigawatts in space. In three and a half years, perhaps 100 gigawatts.

Then, depending on global chip manufacturing and TeraFab's progress, continuing to exceed that scale, reaching 1 terawatt per year, or 1000 gigawatts.

That is twice the electricity consumption of the United States."

Caption: SpaceX's TeraFab design aims to achieve an output of 1 terawatt per year, approximately twice the current electricity consumption of the United States. (Image: terafab.ai)

Comparing Musk to gilded-age industrialists certainly captures some real aspects, but also highlights the differences.

Carnegie built steel.
Vanderbilt built railroads.

Each dominated a sector of the industrial foundation of their time.

Musk is trying to do several sectors at once:

Space, energy, artificial intelligence, robotics, tunneling, brain-machine interfaces, and autonomous vehicles.

And bend them all toward a goal that most consider highly fantastical.

Whether this will work is indeed unknown; many of these may not succeed.

But the attempt itself has no historical precedent and may become a gathering place for another century.

---

The World Opened by SpaceX

Before its retirement in 2011, the space shuttle had a cost of about $54,500 to send 1 kilogram of cargo into orbit.

After Starship matures, Musk expects the cost to drop to $100 per kilogram.

When the cost of entering space drops by more than 500 times, every industry that could theoretically exist in space begins to become economically viable.

There are many such industries.

Caption: The design goal of Starship and Super Heavy is to return to the launch site after flight and be caught by the launch tower, allowing for rapid turnaround and re-launch without refurbishment. (Image: SpaceX)

The closest historical analogy may be the transcontinental railroad in the United States.

Before 1869, traveling from New York to San Francisco took six months by horse-drawn carriage, costing about a year's salary, with a very real risk of death.

After 1869, the journey took only a week.

The railroad itself was an incredible engineering achievement, but the real story is what it opened up:

Sears Roebuck, meatpacking giants like Swift and Armour, Standard Oil, and eventually U.S. Steel—all were born in the railroad boom and further integrated into industrial empires.

If Falcon 9 is the transcontinental railroad of the space age, then Starship may be an upgrade equivalent to that of airplanes.

Railroads opened a continent.

The jet age opened a planet.

Starship will open the solar system.

---

Industrialized Moon

Since humanity began looking up at the Moon, it has always had scientific significance.

Now, it is beginning to have economic significance.

Because it is a complete world made of industrial raw materials.

First, consider how to send things off the Moon.

As mentioned earlier, the Moon has only one-sixth the gravity of Earth and no atmosphere, making mass drivers—the natural way to transport goods from the lunar surface—rather than rockets.

This will fundamentally change the economics of transportation.

Once an orbital infrastructure is established, the marginal cost of transporting finished goods will primarily be determined by electricity, not fuel.

And the power on the Moon comes from sunlight.

A package is thrown from the lunar surface, re-enters the Earth's atmosphere with a heat shield, opens a parachute, and lands at a recovery site.

When the throughput is large enough, the marginal cost will no longer resemble spaceflight but will look more like freight transport.

Next is: What will you manufacture there?

The same lunar regolith can provide the silicon and aluminum needed for solar cells and satellites, and can also serve as the raw material for an entire industrial foundation.

The space revolution of the 2030s and 2040s may look like this:

Automated mining vehicles operating around the clock on the lunar regolith;
Refineries producing aluminum and silicon;
Factories assembling satellites, solar panels, and the chips needed to operate them.

Most industries on Earth have a lunar version waiting to be built.

SpaceX cannot build it all alone.

Those who build "lunar Alcoa," "lunar Caterpillar," and "lunar Union Pacific" will become the giants of the 21st century.

Caption: SpaceX's Starship HLS is the lunar lander designed for NASA's Artemis program, aiming to return humans to the lunar surface after more than 50 years and deliver the foundational modules needed for permanent presence near the lunar south pole. (Image: SpaceX)

---

Computing Power in the Sky

By 2030, the bottleneck for artificial intelligence may not be chips but power.

The obvious response is to build more solar power in Texas or Nevada.

But it hits a wall faster than people think.

1 terawatt of continuous solar power requires about 1% of the land area of the United States.

And new utility interconnection permits typically take a year or longer.

Building Colossus in Memphis required xAI to deploy a fleet of temporary gas turbines, wrestle with state permitting processes, and establish an independent power hub across the state line in Mississippi to bring 1 gigawatt online.

Scaling this to the hundreds of gigawatts needed for AI construction is simply not feasible.

Even the backup gas turbines for solar power have their internal blades and guide vanes backlogged until 2030.

Caption: Baker Hughes Frame 5/2C gas turbine generator. The internal casting blades and guide vanes of such gas turbines are produced by a handful of specialized foundries, all backlogged until 2030. A super-scale data center would require dozens of such devices. (Image: Baker Hughes)

The solution is to move computing power to where the sunlight is already present.

Once Starship flies daily and orbital deployment becomes routine, this will be easier.

And the economics will continue to improve with the cost curves of rocket launches, solar panels, and chips.

SpaceX CFO Bret Johnsen explains:

"We are increasing factory capacity and benefiting from falling silicon costs, so our costs will decrease in the coming years.

If you look at ground solutions, the curve is going in the opposite direction. Everything is getting more expensive: cooling methods, electricity costs are not going down, and land and regulations are becoming more challenging."

A common counterargument comes from those who hear "space data centers" and think it means launching a building the size of Colossus into orbit.

But that is not the case.

SpaceX early investor Gavin Baker said: "It’s probably about the size of a Blackwell rack, with solar wings, possibly 500 feet long on each side. You put it in a sun-synchronous orbit so that the solar panels are always in sunlight.

I've spent a lot of time at Starbase over the years and talked to many SpaceX engineers. I really believe this is the most talented group of engineers on Earth, and they are very confident they have solved this problem."

Caption: AI Sat Mini is designed to capture the sun's energy. (Image: terafab.ai)

In fact, Musk believes that AI Sat Mini is easier to manufacture than Starlink satellites.

He explains:

"You still have some laser links, but you don’t need the extremely complex antennas that are on Starlink satellites.

Compared to that, AI satellites are easier to design.

AI satellites don’t require any magic. A lot of the technology we’ve already developed for Starlink V3 satellites. Compared to what we are already doing, we don’t think this is a particularly difficult problem."

He predicts that within five years, the AI computing power SpaceX launches into orbit each year will exceed the total installed computing power on Earth.

The math here is roughly:

10,000 Starship launches per year, which is more than once an hour around the clock.

By the late 2030s, with the lunar mass drivers coming online, the petawatt threshold will come into view:

That is 1000 times the computing power deployed in 2030, launched into deep space at a rate of one satellite every few minutes.

---

Mars

The Mars trajectory was originally supposed to start this year.

In September 2024, Musk announced that SpaceX would launch five unmanned Starships to Mars during the transfer window in November 2026, carrying Optimus robots to test landing systems, search for ice, and begin building infrastructure for future crewed missions.

He said in May 2025 that the probability of achieving this goal was fifty-fifty.

But earlier this year, the situation changed.

On February 8, Musk posted on X announcing that SpaceX would delay the Mars timeline and shift its immediate focus to self-sustaining cities on the Moon.

The reason is:

Mars launch windows only occur every 26 months, with a flight time of 6 months; whereas the Moon can be reached every 10 days, with a flight time of only 2 days.

He wrote:

"This means we can iterate faster and complete the lunar city compared to the Martian city.

That said, SpaceX will also work to build a Martian city and will start doing so in about five to seven years, but the overwhelming priority is to secure the future of civilization, and the Moon is faster."

On the surface, this seems like a pivot.

But in reality, it is a moment when the path to a million-person city on Mars becomes clearer.

The topic of orbital data centers became clearer by late 2025 to early 2026, giving the Moon a new role.

To achieve petawatt-level orbital computing power, it requires:

Lunar mining, lunar refining, lunar manufacturing of solar panels, radiators, and satellite structures, and launching them into orbit using mass drivers powered by lunar energy.

Such an industrial foundation at scale requires a permanent population, and a permanent population requires a city.

This city can be fully funded by the orbital computing industry while serving as a rehearsal for Mars.

Every problem SpaceX must solve to establish a self-sustaining city on Mars will first be encountered in the lunar city:

• Radiation shielding;
• Life support;
• In-situ resource utilization;
• Governance of a permanent population off Earth;
• Supply chains crossing gravity wells.

Building a lunar city will teach SpaceX how to build a Martian city through a much faster iterative cycle.

The first unmanned lunar landing demonstration is targeted for as early as 2027.

According to Musk's public timeline, the lunar city will follow in less than ten years.

The lunar manufacturing of mass drivers, lunar industrial construction, and orbital computing infrastructure will all be initiated in parallel.

Then, Mars.

But the hardest part is not transporting people.

It is building the Martian infrastructure to absorb those people.

The lunar rehearsal will help.

Optimus will also help.

Musk repeatedly mentioned in his May 2025 Starbase Mars speech that early unmanned Starships would carry Optimus robots to search for resources and begin building infrastructure for human arrival.

The company is building a production line capable of producing 1 million units per year in Fremont and another line capable of producing 10 million units per year in Giga Texas.

These robots are still in early production stages and have not yet completed meaningful practical work in Tesla factories.

But the capacity coming online in the next two to three years will be crucial for the initial self-sustaining Martian base.

Caption: SpaceX rendering: Optimus robots working on Mars, replicating the famous photo "Lunch atop a Skyscraper" from the construction of Rockefeller Center in 1932. (Image: SpaceX)

---

Conscious Sun

The mission statement adopted by SpaceX after absorbing xAI in February is:

Scale to give birth to a conscious sun, to understand the universe and extend the light of consciousness to the stars.

How you interpret this statement depends on your perspective.

It is either the most absurd thing ever written on a mission page by a serious company,
or the most honest.

We believe it is the latter.

If you squint at the organizational structure, SpaceX is a launch service provider with an internet subsidiary and a recently acquired AI lab.

If you squint at the technology roadmap, it is the only company on Earth assembling a complete pre-scarcity transformation stack.

If you squint at the mission statement, it is a founder with operational capabilities among the most significant of our time, earnestly trying to push humanity through that bottleneck.

On the other side of the bottleneck are two possibilities:

One is that we become an interstellar species, sharing the universe with the intelligent machines we build ourselves;

The other is that we are merely a footnote on some rocky planet that has not completed its leap.

When the first child born on Mars asks their parents: Why are we here? Starship may have been flying routinely for thirty years.

In the factory around the corner, Optimus robots are working, running descendants of Grok that have been self-improving for twenty years.

The computing power sustaining her city comes from space data centers.

These data centers are manufactured from lunar regolith by other robots and launched by a mass driver. That mass driver has been flinging satellites into deep space at a rate of one every few minutes for nearly a generation.

Her parents arrived on Mars aboard a vehicle named after a starship from Iain M. Banks' novels.

Because at some point in the early 21st century, a teenager who read those books decided to spend a lifetime making them a reality.

Banks understood those who choose to go to Mars.

"The Culture" is paradise, but his most interesting characters are often those who leave paradise.

The Culture solved scarcity; what remains is humanity's desire for the difficult journey.

Even if paradise is next door, the frontier remains the place where meaning dwells.

Musk said the early Martian colonists' recruitment slogan would be Shackleton-esque.

It comes from the famous recruitment ad for the 1914 Antarctic expedition:

"Men wanted for a hazardous journey. Low wages, bitter cold, long months of darkness, constant danger, and the possibility of not returning safely. Successful candidates will gain honor and recognition."

This ad was almost certainly fabricated by later generations.

But the story has been told for a hundred years because it captures some truth about the choice of those who set out.

Why would anyone find this appealing?

Musk said: "Life cannot just be solving one miserable problem after another. There must be something that inspires you, that makes you wake up in the morning and be glad to be part of humanity. Earth is humanity's cradle, but you cannot stay in the cradle forever. It is time to go out and become a civilization that reaches for the stars, to walk among the stars and expand the scope and scale of human consciousness. I find that incredibly exciting. It makes me glad to be alive. I hope you feel the same way."

Caption: Starman, a mannequin in a SpaceX spacesuit, sits at the wheel of Elon Musk's private Tesla Roadster, orbiting the Sun. This car was the payload for the maiden test flight of Falcon Heavy, launched on February 8, 2018. Its current orbit will pass near Mars approximately once every Earth year for the next million years. (Image: SpaceX)

Disclaimer

This material is for educational purposes only and does not constitute investment advice or an offer of investment advisory services.

This material should not be the basis for any investment decision.

a16z has invested in SpaceX through its managed funds and therefore has a financial interest in the company's performance and future prospects.

In particular, if SpaceX's value increases, a16z will benefit from it; as a shareholder in the company, a16z's funds will also receive any customary dividend payments.

However, SpaceX has not compensated a16z for this material.