Elon Musk Says SpaceX’s $28.5 Trillion Market Estimate May Be Far Too Low
Elon Musk believes SpaceX may have dramatically underestimated the size of its future business opportunity. The company previously outlined a massive $28.5 trillion total addressable market, or TAM, in filings tied to its public market ambitions, but Musk now says the real figure could be “vastly higher.”
That statement adds another layer of ambition to SpaceX’s long-term strategy. The company began as a rocket launch provider, then transformed into a global satellite internet operator through Starlink. Now, SpaceX is increasingly positioning itself as a major player in artificial intelligence infrastructure, a market that could reshape the company’s future far beyond space launches and broadband connectivity.
A key part of this vision is AI compute. SpaceX has described an enormous market opportunity tied to enterprise artificial intelligence, consumer AI services, digital advertising, and global connectivity. In its previous estimate, the company suggested that most of its projected $28.5 trillion opportunity would come from AI-related services.
According to the figures outlined by the company, roughly $26.5 trillion of the market estimate was tied to artificial intelligence. Within that category, enterprise AI applications were expected to account for approximately $22.7 trillion. Consumer AI subscriptions were estimated at around $760 billion, while digital advertising was projected at about $600 billion. Starlink and connectivity services made up much of the remaining opportunity, with SpaceX estimating a connectivity market of around $1.6 trillion.
Even with those enormous numbers, Musk now believes SpaceX may have been too conservative. In a post on X, he responded to discussion of the company’s market estimate by saying, “We were wrong. It is vastly higher than that.”
The comment suggests that Musk sees SpaceX’s future not just as a space transportation and internet company, but as a foundational AI infrastructure provider. That aligns with his broader push into large-scale computing, data centers, and advanced AI systems.
One of the clearest examples of this strategy is the Colossus data center project. Musk has been working to bring hundreds of thousands of high-performance AI chips online to support large-scale artificial intelligence workloads. The Colossus 2 facility in Memphis is expected to use as many as 660,000 of NVIDIA’s latest Blackwell AI GPUs by the end of the year. These would add to the hundreds of thousands of Blackwell GB300 and GB200 chips already associated with the facility.
The scale of that buildout highlights just how aggressively Musk is moving into AI infrastructure. Advanced GPUs are among the most important resources in the AI industry, powering the training and deployment of large language models, generative AI platforms, autonomous systems, enterprise automation tools, and other compute-heavy applications.
But SpaceX’s AI ambitions may extend even further than Earth-based data centers. The company has also discussed Starmind, a concept that appears to apply the Starlink model to artificial intelligence computing. Just as Starlink relies on thousands of satellites to deliver global internet coverage, Starmind would use satellites to provide distributed AI compute capacity from orbit.
The goal is bold: SpaceX wants future Starmind satellites to deliver significant average compute power, with comparisons being made to NVIDIA’s GB300 Blackwell platform. Musk has indicated that launches could begin as early as late 2027, and SpaceX is reportedly working with NVIDIA to adapt advanced computing platforms for space-based infrastructure.
If successful, Starmind could become a major step toward orbital AI computing. Such a system could potentially serve customers that need low-latency, distributed, resilient, or globally available compute power. It could also create an entirely new market category, combining satellite networks, artificial intelligence, and cloud computing into one platform.
SpaceX’s expanding vision comes at a time when the AI infrastructure market is attracting intense attention across the technology industry. Major chipmakers, cloud companies, and hyperscale data center operators are all racing to secure more computing capacity. AMD has estimated a $1.5 trillion opportunity for data center AI accelerators, while NVIDIA has pointed to trillions of dollars in demand from hyperscale customers and AI infrastructure buildouts.
Against that backdrop, Musk’s claim that SpaceX’s $28.5 trillion market estimate is too low reflects the possibility that AI demand could exceed even the most aggressive forecasts. As businesses, governments, developers, and consumers increasingly rely on AI tools, the need for compute power, connectivity, and scalable infrastructure may continue to surge.
For SpaceX, this could mean a future where rockets, satellites, Starlink internet, AI data centers, and orbital compute systems all work together as part of a much larger ecosystem. The company’s long-term opportunity may no longer be limited to launching payloads or providing broadband access. Instead, SpaceX appears to be aiming at the backbone of the AI economy itself.
Musk’s latest remark does not provide a revised number, but it does send a clear message: SpaceX believes its addressable market could be far larger than previously stated. If the company can successfully combine its launch capabilities, satellite network, AI infrastructure plans, and computing partnerships, its role in the future of artificial intelligence could become much bigger than many observers expected.
