Here’s what’s on our plate today:

• 🚀 SpaceX and xAI merger and orbital AI data centers.

• 🧠 Google AI warnings, Alibaba’s upgrade, and India’s AI diplomacy.

• 💡 Roko’s Pro Tip on sanity-checking trillion-dollar AI plays.

• 📊 Monday poll on whether Musk’s mega-merger actually works?

Let’s dive in. No floaties needed…

The Laboratory

Making sense of SpaceX’s $1.25T merger with xAI

Across modern economic history, several business leaders have attempted to build companies capable of dominating their industries and leaving a lasting impact on human society. However, achieving such success takes more than determination; it requires favorable technological advancements, sustained market trust, access to capital, regulatory tolerance, and the ability to control or influence the critical infrastructure on which competitors depend.

In the age of artificial intelligence, the pursuit of dominance introduces an additional constraint rarely encountered at this scale: physics. AI firms are no longer limited solely by algorithms or data. They are bound by energy availability, cooling capacity, and the environmental realities of running compute-intensive systems. Unlike earlier technological revolutions that unfolded over decades, AI capabilities are advancing so quickly that infrastructure must expand to keep up, often straining power grids and supply chains.

Modern AI data centers consume extraordinary amounts of electricity, tying computational growth to regional energy systems. Companies, therefore, face a strategic dilemma. They must either compete for finite grid resources, invest in alternative energy arrangements such as nuclear partnerships, or explore entirely new models of power generation and distribution. In this area, Elon Musk stands out as a businessman pushing at the very limits of what was thought possible. And to do that, he is now making the strategic move of uniting three of his most well-known enterprises to address some of AI’s most complex physical limitations.

SpaceX, xAI, and vertical integration

Through this unification, Musk, long known for his affinity for launching ambitious projects, is now on a path to build AI data centers in space. The merger of SpaceX and xAI is positioning the companies to overcome significant hurdles through a tightly integrated ecosystem of rockets, satellites, and AI systems, enabling AI infrastructure to extend beyond Earth.

According to Musk, the merger between SpaceX and xAI, a smaller firm known for its Grok chatbot, establishes the most ambitious, vertically integrated innovation engine on (and off) Earth, encompassing AI, rockets, space-based internet, and the X social media platform.

Although the exact terms of the deal between Musk’s companies are not known, a BBC report valued xAI at $125B (£91B) and SpaceX at $1T, making it the most valuable private company ever.

Additionally, Tesla, Musk’s electric vehicle manufacturer and the source of most of his liquid wealth, said last week that it was investing about $2B in xAI.

According to Reuters, the ambition behind the newly combined entity is already evident. SpaceX has requested authorization to deploy as many as 1M solar-powered satellites designed to operate as orbital data centers, a scale that would surpass any space-based infrastructure currently deployed or even seriously proposed.

Additionally, in documents submitted to the Federal Communications Commission, the company outlined plans for a solar-powered network built around optical inter-satellite links, which it referred to as an orbital data center system. The filing, however, did not specify how many Starship launches are required to expand the network to a commercially viable level.

While these appear to be concrete steps on paper, it will take time for these ambitions to translate into practice. In the meantime, the company has to navigate regulatory and real-world challenges before it can begin building data centers in space.

The logic of the triangular merger

To begin with, Musk structured the SpaceX xAI deal as a triangular merger, rather than a full combination of the two companies. xAI remains a wholly owned subsidiary of SpaceX, preserving its separate legal and financial identity. According to Reuters, the approach achieves three objectives: it defers capital gains taxes for xAI shareholders, avoids triggering debt covenants that could have required immediate repayment of billions in borrowings, and insulates SpaceX from xAI’s legal liabilities.

That insulation is far from trivial; xAI inherited $12B in debt when it acquired X in 2025 and has added at least $5B more since. The subsidiary structure means SpaceX does not assume responsibility for that debt. It also means the regulatory probes circling xAI on multiple continents do not, on paper, attach to SpaceX’s balance sheet.

Contagion risk across borders

The phrase “on paper” carries significant weight here. Columbia Law professor Eric Talley told CNBC that regulators in some jurisdictions may still require the parent company to remain in good standing overall to operate within their borders. For Starlink, which serves customers across dozens of countries, enforcement actions against xAI in Europe or Asia could pose tangible risks to SpaceX’s satellite business.

Those risks are not merely theoretical. Since December 2025, xAI has been embroiled in a deepfake crisis that has put regulators worldwide on alert.

The crisis began when, despite warnings from experts, a Grok update enabled mass production of nonconsensual deepfakes, including images of children.

It prompted the European Commission to open an investigation under the Digital Services Act against the company. Additionally, French prosecutors raided X’s Paris offices and ordered Musk and CEO Linda Yaccarino to face mandatory hearings in April on charges including complicity in the distribution of child pornography. Three countries blocked Grok entirely, and a coalition of 35 U.S. State Attorneys General demanded that xAI halt the creation of nonconsensual images. The California AG issued a cease-and-desist under the state’s new deepfake law. The company is also facing a class action lawsuit in federal court. All this is now part of SpaceX’s IPO story.

Human capital under stress

To make matters even more difficult for Musk, xAI is also struggling to retain the talent needed to realize its vision.

As of this week, six of xAI’s 12 original co-founders have departed the company. Jimmy Ba, credited with research that shaped the Grok 4 models and responsible for the safety research agenda, announced his exit on 10 February. Tony Wu left the day before. They followed Igor Babuschkin, Kyle Kosic, Christian Szegedy, and Greg Yang.

Musk described the exits as part of the company’s evolution. During an all-hands meeting, he likened the departures to natural selection, suggesting that some employees were better suited to a startup’s early phase. Later, on X, he clarified that the changes were involuntary, noting that XAI had been reorganized to move faster, which meant letting certain employees go.

However, TechCrunch reported that at least three departing engineers are starting new ventures together, a pattern that suggests something more than a standard reorg. When the safety lead leaves during a safety scandal, and the founding researchers leave together to build something new, the signal is not just operational; it is directional.

In frontier AI, where world-class researchers can choose from any lab on the planet, the ability to recruit and retain talent is as important as any balance sheet metric. Losing half the founding team while under investigation for safety failures, in a market where competitors like Anthropic and OpenAI are actively courting safety-conscious researchers, is a compounding problem that no triangular merger can contain.

What investors will really price

Musk may be seeking to establish an entity capable of addressing some of the AI industry’s most pressing challenges by building data centers in space. However, the path from here to there runs through courtrooms, regulatory agencies, and an IPO road show, where every one of these risks will be scrutinized by institutional investors with their own fiduciary obligations.

So while the triangular merger protects the balance sheet, it does not protect the narrative. And in the largest tech IPO in history, the narrative is almost as important as the product.

The triangular merger may reduce financial exposure, but it cannot isolate reputational and regulatory risk. As SpaceX moves toward what could be the largest technology IPO ever, investors will evaluate not only rockets, satellites, and AI models but also the company’s broader strategy. They will evaluate governance, legal uncertainty, and the durability of market trust. History suggests that scale and ambition alone rarely settle those questions.

Roko Pro Tip

Bite-Sized Brains

• Google’s AI overviews under fire: A Guardian investigation says Google watered down safety warnings while pushing error-prone AI Overviews into core Search, potentially putting users at risk. 

• Alibaba rushes out a DeepSeek-era upgrade: Alibaba is touting a major Tongyi/Qwen model refresh, positioned to keep pace with DeepSeek’s reasoning push and to reassure investors that it can still compete on quality, not just scale. 

• India turns Delhi into an AI power stage: The India AI Impact Summit brings global tech CEOs and leaders to New Delhi as Modi pitches the country as a key hub for “responsible” AI and big-ticket investment. 

Monday Poll

Meme of the Day

Rate This Edition