SpaceX’s Ambition: From Rockets to AI—What Game Is Musk Playing Next?

Lee, Golden Finance

In May 2026, the tech world was blindsided by two pieces of news. First, Elon Musk—who has long slammed Anthropic as “inhumane” and “evil” and claimed it “hates Western civilization”—suddenly announced a major computing power partnership with the company, a former nemesis. At the same time, on the X platform, he casually added, “This time, the partnership didn’t trigger my evil detector.” This slightly playful remark appears to be Musk joking about his own attitude. In reality, it’s an important signal for consolidating his business landscape. A collaboration that looks like a “slap in the face” has never been an impulsive compromise; rather, it is a key move in the multi-year layout he has been planning. Musk’s AI strategy has never been an isolated story separate from SpaceX or Tesla. Instead, it is the core puzzle connecting all of his business empires— and SpaceX is at the center of this grand chessboard.

SpaceX’s Three Moves

Few people truly understand that from the moment SpaceX was founded, it has never been merely a rocket company. The ultimate mission Musk set for it is “to make humanity a multiplanetary species.” Behind that mission lies a clear, three-step progression—each step plants a seed for the final AI plan, building on the previous one in a layered, tightly connected way.

The first step is rockets—the “entry ticket” that breaks the space monopoly. Before SpaceX appeared, space exploration was the exclusive domain of national space agencies. The high-cost, low-efficiency model made commercial space exploration seem like a fantasy. With the reusable Falcon 9 rocket, Musk completely upended industry rules, compressing space transportation costs to a fraction of what they used to be. He even planned that with the Starship V3 rocket, costs would be lowered further to 100 dollars per kilogram—just 1% of the current industry level. The core of this step is to secure “space access rights.” Only by being able to send equipment into space cheaply and efficiently can all subsequent space deployments become feasible. Like playing chess, rockets are Musk’s “pawns”—seemingly insignificant, yet capable of gradually opening up the path deeper into the board and clearing obstacles for the placement of later pieces.

The second step is Starlink—building a “neural network” for space communications. Having rockets as the transport means, Musk immediately launched the Starlink project, aiming to deploy tens of thousands of satellites into low Earth orbit and build a globally seamless satellite communication network. By the end of 2025, the number of Starlink satellites in orbit had exceeded 9000. The service covered 150 countries and regions, and global active subscribers reached 7.65 million. Starlink not only enabled breakthroughs in direct mobile connectivity, but was also used by the U.S. Department of Defense for military surveillance. In this way, Musk built a business closed loop of “rocket launch—satellite networking—communication services.” In this step, what Musk wants to create is a space version of the “internet.” Starlink can provide communications services to Earth and, at the same time, become the core carrier for future space data transmission. Like a neural network, it connects every node on Earth and in space, providing indispensable communication support for future space data centers.

The third step is space data centers and orbital AI computing—unlocking the “ultimate form” of AI. When ground-based computing power increasingly runs into bottlenecks such as power shortages and heat dissipation difficulties, Musk had already turned his attention to space. In January 2026, SpaceX formally applied to the Federal Communications Commission (FCC) to launch and operate 1 million low-orbit AI satellites, planning to build the world’s first orbital AI data center network encircling the Earth. This scale equals 100 times the current total number of satellites in orbit globally. The natural cooling advantages of space’s vacuum, ultra-low-temperature environment allow data centers’ PUE to be as low as 1.05, far better than the 1.2–1.4 levels typical of ground data centers. Meanwhile, 24 hours of uninterrupted sunlight enables solar power generation efficiency to reach 7–10 times that of the ground. This allows the system to completely escape constraints from terrestrial power grids, land resources, and heat dissipation. This partnership with Anthropic takes the idea into reality: Anthropic will lease all computing power from SpaceX’s Colossus 1 data center in Memphis, Tennessee. The facility is supported by more than 220,000 Nvidia AI chips and can provide a computing capacity of 300 megawatts. At the same time, the two sides also plan to collaborate on developing “multi-gigawatt-level orbital AI computing.” From rockets to Starlink, and then to space data centers and orbital computing, SpaceX’s three-step moves are, in essence, a leap from “space transportation” to “space infrastructure,” and then to “space computing.” All of it is designed to carry Musk’s AI ambitions.

Dissolving xAI

On the same day as the Anthropic partnership announcement, Musk revealed another major development: xAI will no longer exist as an independent company. Going forward, it will operate under the identity of “SpaceXAI” and become the AI product division under SpaceX. Grok and X (formerly Twitter) will also be folded into this system. The decision shocked the public—after all, since xAI was founded in 2023, in just three years its valuation has surpassed $250 billion, and it released the much-discussed Grok robot. So why would Musk suddenly dissolve an independent company with such promising prospects? The answer is simple: for SpaceX’s IPO, to maximize valuation, and to complete the final consolidation of the business layout.

First comes brand integration to remove business barriers. During xAI’s period of independent operations, it remained separated from SpaceX and Tesla. xAI needed computing power but had to rely on SpaceX’s infrastructure. SpaceX needed AI technology to empower space exploration but had to coordinate across companies with xAI. Tesla’s autonomous driving AI technology also had a resource fragmentation problem relative to xAI’s general AI R&D. Merging xAI into SpaceX and forming SpaceXAI can fully connect AI R&D, space computing, satellite communications, and rocket launches—creating a “AI + space” synergy. SpaceX’s space infrastructure provides scenarios and computing power for AI. In return, AI technology optimizes the efficiency of rocket launches, Starlink operations, and space exploration, producing an effect where “1+1 is greater than 2.” This kind of integration is like gathering scattered chess pieces—so each piece can deliver its maximum value instead of fighting on its own.

The more fundamental reason is to boost SpaceX’s IPO valuation. According to market reports, SpaceX plans to start its IPO in late June 2026. The target valuation is as high as $1.75–2 trillion, and it plans to raise $75 billion—far exceeding Saudi Aramco’s earlier IPO record of $29 billion, with a strong chance of becoming the largest IPO in history. Against this backdrop, xAI’s continued independent existence would actually divert market expectations for SpaceX’s valuation. xAI’s AI business is a core asset with high growth and high valuation. By integrating it into SpaceX, SpaceX can upgrade from a “space company” to a “space + AI” technology giant, completely changing the valuation logic. Previously, the combined entity after the SpaceX and xAI merger had an implied valuation of $1.25 trillion. With xAI officially renamed SpaceXAI and added to the computing partnership with Anthropic, SpaceX’s valuation is expected to push toward a $2 trillion mark. For Musk, dissolving xAI is not abandoning the AI business—it is placing this most valuable “chess piece” into the position that can raise the overall value the most, thereby maximizing capital value.

Capital Operations Before SpaceX’s $2 Trillion Valuation IPO

A valuation of $2 trillion means SpaceX’s IPO will become a once-in-a-generation feast that shakes global capital markets. Before this feast begins, Musk had already teamed up with Wall Street and carried out a series of precise capital moves to pave the way for the IPO. The core of these moves is to mobilize market funds, strengthen valuation logic, and hedge potential risks to ensure the IPO lands smoothly.

The most attention-grabbing is that Wall Street institutions are discussing selling tech stocks to free up funds for subscribing to SpaceX shares. Reports from Reuters, Bloomberg, and others indicate that as SpaceX’s listing comes closer, fund managers are actively researching how to reserve capital for subscriptions. Tony Wang, who manages the $12 billion T. Rowe Price Technology Fund, stated that his decision is made in the context of “Mag 7” (including seven tech giants such as Apple, Microsoft, Nvidia, and Tesla). The key is weighing which offers the “largest risk-adjusted return.” Some fund managers also believe that many Tesla investors will sell part of their Tesla holdings and instead subscribe to SpaceX.

Behind this fund shuffling are multiple real-world factors. On one hand, industry concentration limits force institutions to rebalance their holdings. Currently, tech stocks account for about two-thirds of growth-oriented funds’ holdings. After listing, SpaceX will be classified in the tech or aerospace sector, and its inclusion will further raise the share of the portfolio. To avoid violating internal risk controls or sector allocation rules, some funds that want to buy SpaceX must reduce their existing tech stock holdings accordingly. On the other hand, tax-loss harvesting has become an important incentive. Some tech stocks have recently declined. For example, Microsoft’s stock is down by about 25% since its October 2025 peak. Selling these loss-making stocks can offset taxes and, at the same time, raise funds to subscribe to new SpaceX shares—killing two birds with one stone. Additionally, “Mag 7,” with its large market capitalization and strong liquidity, has become the best option for institutions to quickly raise cash.

Beyond reallocating funds, Musk is also strengthening SpaceX’s valuation logic through a series of strategic moves. First, expanding the retail allocation: he plans to allocate up to 30% of IPO shares to retail investors worldwide, far above the usual 5%–10% range. This can attract more capital into the market, increase overall attention, and push the stock price higher. Second, pushing for index inclusion. S&P Dow Jones is considering reforming the rules so that the time requirement to be added to the S&P 500 for new companies would be shortened from 12 months to 6 months. If the reform passes, SpaceX could attract passive index-tracking funds of about $24 trillion to buy shares automatically within six months after listing, providing long-term support for the share price.

But behind this capital feast, risks are also lurking. Jay Ritter, a professor at the University of Florida known as the “IPO Guru,” warned that if SpaceX goes public with a $2 trillion valuation, he would choose to short it. Such a valuation requires Starlink to maintain an astonishing 60% average annual growth, and it depends on all high-risk projects—including the successful commercialization of Starship rockets and space AI data centers—being completed smoothly. SpaceX also acknowledges in its IPO filing documents that its grand plans rely on “unproven technology.” Even so, enthusiasm on Wall Street remains high. After all, in this largest IPO in history, whoever can successfully “get on board” stands to gain outsized returns. Musk, as the mastermind behind this capital game, is the one orchestrating it all.

From Tesla AI to xAI to SpaceXAI

Dissolving xAI and establishing SpaceXAI may seem like a business adjustment, but in reality it is the final consolidation of Musk’s AI empire. It brings the AI resources scattered across Tesla, xAI, and SpaceX into one cohesive whole. Looking back at Musk’s AI layout, it becomes clear that this was never a spur-of-the-moment adventure. It is a long-term strategy spanning from automobiles to space, from local plans to a global vision. Each step is tightly connected, ultimately building an AI empire spanning multiple domains.

Musk’s AI layout began with Tesla. Long before Tesla launched autonomous driving features, Musk realized that AI would be the core competitive advantage in transportation and energy in the future. Tesla’s AI team focuses on developing algorithms for autonomous driving. By training on massive vehicle driving data, it produces more accurate and safer autonomous driving models. In this process, the accumulated AI training experience and data processing capabilities became the foundation for Musk’s subsequent AI plans. In a sense, Tesla AI has been the “testbed” for Musk’s AI empire: it validates the application value of AI in real-world industries and also helps cultivate key talent for what comes next, including xAI.

In 2023, the founding of xAI marked Musk’s AI strategy shifting from “vertical domains” to “general domains.” xAI’s initial team consisted of 11 top AI researchers, including Igor Babuschkin, a former chief engineer at Google DeepMind. Its core goal was to develop an “ultimate truth-seeking” general artificial intelligence (AGI). The Grok robot it launched gained rapid attention in the AI community due to its unique interaction approach and strong reasoning capability. In 2025, xAI completed its acquisition of the X platform (formerly Twitter), further enriching its data resources. The X platform’s massive user data became valuable material for training general AI models, helping xAI’s valuation quickly climb to $80 billion. Then in 2026, SpaceX acquired xAI via an all-stock transaction, raising the valuation to $250 billion.

The establishment of SpaceXAI signals that Musk’s AI empire is now fully formed, achieving full-scenario coverage of “AI + space + automobiles.” Today, SpaceXAI integrates xAI’s general AI R&D capabilities, Tesla’s autonomous driving AI technology, and SpaceX’s space computing infrastructure to create a complete AI ecosystem. In the space domain, AI is used to optimize rocket launch trajectories, schedule Starlink satellites, operate space data centers, and even develop orbital AI computing. In the automotive domain, autonomous driving AI technology continues to iterate, pushing Tesla toward a transition into a “smart mobility” terminal. In the general AI domain, the Grok robot continues to be upgraded, combining X platform user data to build AI products that are more closely aligned with user needs. More importantly, SpaceX’s orbital computing power can not only meet the needs of its own AI R&D, but can also monetize through rental. Just like the partnership with Anthropic: it solves the problem of idle computing capacity while adding a steady source of revenue to the AI empire.

From Tesla AI to xAI to SpaceXAI, Musk’s AI empire has never been an isolated set of assembled parts. Rather, it is an integrated whole where everything enables the other parts and develops in synergy. Tesla provides AI with application scenarios and data. xAI provides general AI technology. SpaceX provides the ultimate computing platform. Together, they form the core competitiveness of Musk’s business landscape— and all of it revolves around a single ultimate goal: using AI and space technologies to reshape humanity’s future way of living.

Is He a Genius or a Con Artist?

Every step in Musk’s planning comes with intense controversy. Some people hail him as the “Iron Man of Silicon Valley,” praising him as a genius who changes the world through extreme innovation and pushes human technology forward. Others denounce him as a “con artist who’s good at telling stories,” arguing that all his grand blueprints are nothing more than hype designed to inflate valuations and harvest capital. This debate has never stopped. And Musk’s complexity lies in the fact that he is both a genius and also hard to escape the games played with capital, making it difficult to define him with a single label.

But doubts about him are not without basis. Some believe he is good at “selling big dreams.” Many of his ambitious plans have serious technical bottlenecks and implementation difficulties. For example, SpaceX’s million-satellite AI plan faces numerous challenges such as uncertainty around FCC approval and competition for orbital resources. The concept of space data centers also depends on the successful commercialization of Starship rockets, a technology that has not yet fully matured. A $2 trillion IPO valuation corresponds to a price-to-sales multiple of 117 based on 2025 revenue—far above mature giants like Apple and Microsoft—meaning the bubble risk is severe. Critics have also pointed out that Musk’s business empire has consistently been accompanied by governance controversies and ethical issues. Tesla has been reported for financial fraud, with 27% of delivered vehicles allegedly involving “false accounting.” SpaceX has been reported to have secretly sold Starlink equipment to Iran and the Myanmar military government, potentially violating U.S. export controls. He also enforces “absolute control” within his companies, leaving employees facing a high-pressure work environment. In addition, many of his predictions have been questioned for being overly optimistic. Chinese economist Li Daokui, for instance, once said that Musk’s understanding of the laws governing how human society operates is still immature. Without institutional reforms to support technological progress, it would only worsen income and wealth inequality rather than enabling everyone to achieve high incomes.

Looking back at Musk’s layout—from rockets to Starlink, from xAI to SpaceXAI, and from Tesla to space data centers—this chess game he has played is grand and complex. The core of this game has never been just a single rocket technology or isolated AI R&D. It is about using AI to empower space, and using space to carry AI, ultimately building a commercial ecosystem that covers humanity’s future life. Whether he is a genius or a con artist, what cannot be denied is that he is reshaping the landscape of the space and AI industries in his own way, pushing humanity toward a more distant future. SpaceX’s IPO may be the key move in this grand chessboard. As for whether the final pieces can be played into a winning position, only time will tell.