Why Did He Dodge the Tesla-SpaceX Merger Question?

The question was simple. The answer was not.

The headline question is still the right one: Why did he dodge the Tesla-SpaceX merger question? In the latest discussion around Morgan Stanley’s physical-AI thesis, Adam Jonas was pressed on whether Tesla and SpaceX should be treated as future merger candidates. He did not offer a base-case probability, a timetable, an exchange ratio or a transaction structure. Instead, he described the relationship as increasingly “deterministic,” with Tesla operating as a manufacturing and data-collection layer and SpaceX serving as a connectivity and AI-infrastructure layer [1].

That answer was not a denial. It was also not a confirmation. It was a disciplined refusal to turn an industrial thesis into a deal prediction.

The distinction matters. A merger question asks: will two companies become one legal entity? Jonas’s answer asked a different question: are Tesla and SpaceX already converging operationally around the same physical-AI stack? In his framing, the important development is not whether lawyers draft a combination agreement tomorrow. It is whether robots, vehicles, satellites, chips, batteries, data centers and edge computers are becoming parts of a single economic machine [1].

Physical AI, not just corporate drama

Morgan Stanley’s thesis is that the next large AI wave may move from screens and cloud chatbots into the physical world. Jonas argues that combining AI with robotics could multiply global output by 8 to 10 times before today’s workers retire, against a global GDP level described around $180 trillion in 2025 [1].

That estimate is enormous, and it should be treated as a long-range scenario rather than a near-term forecast. But it explains why the merger question keeps resurfacing. If the big prize is not merely electric vehicles or rockets, but the automation of physical labor, then Tesla and SpaceX look less like unrelated Elon Musk companies and more like adjacent layers in one system.

Tesla brings factories, vehicles, battery systems, autonomy data and a humanoid-robot program. SpaceX brings launch capability, satellite connectivity, advanced manufacturing and a path toward distributed compute. In Jonas’s telling, the companies are both trying to convert energy into intelligence at scale [1].

That phrase can sound abstract, but the mechanics are concrete. Physical AI needs power, chips, motors, actuators, sensors, communications links, local inference and fleets of machines that can collect data from the real world. A robot that cannot connect, compute or move cheaply is not a scalable economic agent. A satellite network without useful machine endpoints is also incomplete. The strategic overlap is the reason the merger question has become difficult to dismiss, even when no formal deal is on the table.

Why a direct answer would be risky

Jonas had several reasons to avoid a direct merger call. First, a transaction between public companies is not just an analyst thought experiment. It would involve boards, valuation, shareholder approval, governance, minority-investor protections and regulatory review. A casual probability could be misread as inside knowledge or as a recommendation on a deal that does not exist.

Second, a formal combination may be less important than operational integration. If Tesla and SpaceX can share technology, infrastructure, purchasing power and strategic direction without merging, investors may still get many of the industrial benefits while avoiding the complexity of a mega-cap stock transaction.

Third, the governance question is sensitive. Tesla shareholders would want to know whether any deal fairly values Tesla’s autonomy, energy and robotics upside. SpaceX investors would want to know whether they are giving up a high-growth aerospace and connectivity platform too cheaply. A merger could be industrially logical and still be financially controversial.

That is why “deterministic relationship” is such a careful phrase. It says the companies’ missions are converging, but it does not say the equity must be combined.

The robots are bigger than humanoids

One important part of Jonas’s argument is that humanoid robots are not the whole story. The latest summaries of his comments emphasize that humanoids may reach very large numbers in some scenarios, but they are only one form factor among many: autonomous cars, low-altitude machines, industrial robots, terrestrial mobile systems and other embedded-AI devices [1].

That matters for Tesla. Optimus receives attention because it is visually compelling and fits the popular idea of a general-purpose robot. But the real physical-AI economy could be broader and less theatrical. A warehouse robot, a driverless car, a drone, a factory inspection machine or an agricultural rover may not look human, yet each could perform economically valuable work.

It also matters for SpaceX. If the world contains tens of billions of connected machines, those machines will require communications coverage and distributed intelligence. VoiceTube’s posting of the Bloomberg segment highlighted Jonas’s view that more inference will need to happen inside machines themselves, rather than relying only on remote data centers [2]. That pushes the discussion toward edge AI: local compute that can react quickly, save bandwidth and keep machines useful even when connectivity is imperfect.

Edge inference changes the architecture

The move from cloud AI to physical AI changes the bottlenecks. A chatbot can wait for a data center. A moving robot cannot always do that. Machines operating in streets, homes, factories or logistics yards need rapid perception, planning and control. Jonas’s argument that inference must move closer to the machine is therefore central to the Tesla-SpaceX discussion [2].

Tesla’s possible role is obvious: cars and robots can become mobile inference nodes. SpaceX’s possible role is equally clear: Starlink-style connectivity can link those nodes across geographies where terrestrial networks are limited or unreliable. The merger question arises because the two layers reinforce each other. Tesla’s machines become more useful if they are always connected. SpaceX’s network becomes more valuable if billions of machines need data pipes.

Yet that still does not prove a merger. It proves interdependence. Interdependence can produce joint ventures, supply agreements, shared platforms or cross-company engineering projects. A merger is only one possible end state.

The hardware bottleneck

The physical-AI thesis also runs into a stubborn reality: robots are made of parts. FrontierNews.ai’s recent actuator analysis underscores that actuators are the “muscles and joints” of robots, and that a single Tesla Optimus unit may require more than 30 of them [3]. The same report says actuators can represent about 55% of a humanoid robot’s bill of materials, turning supply chains into a central constraint rather than a footnote [3].

This is where Jonas’s broad answer becomes more persuasive. If the future is tens of billions of machines, the question is not only who writes the best AI model. It is who can manufacture, power, connect, repair and finance the physical fleet. That is a Tesla question. It is also a SpaceX question. And it is a supply-chain question that reaches far beyond both companies.

FrontierNews.ai also notes Morgan Stanley’s view that dependence on Chinese manufacturing ecosystems, rare earths and component supply cannot be unwound quickly [3]. That point complicates the clean narrative of American reshoring. Physical AI may create demand for U.S. manufacturing, but it may also require a delicate, politically uncomfortable level of global supply-chain cooperation.

So, why did he dodge?

He dodged because the merger question is narrower than the story. A yes-or-no answer would have turned a structural thesis into a market headline. Jonas appears to be saying that the more important fact is already visible: Tesla and SpaceX occupy complementary positions in the physical-AI stack.

That does not mean a merger is inevitable. It means investors should watch the practical links: shared infrastructure, chip strategy, robotics, autonomy, Starlink connectivity, energy storage, edge inference and manufacturing capacity. If those links deepen, the market may continue to price in the possibility of a corporate combination even without an announcement.

The best interpretation is that Jonas avoided the legal question and answered the industrial one. Tesla and SpaceX may or may not merge. But in the physical-AI race Morgan Stanley describes, they are already being analyzed less as separate stories and more as connected engines in the same machine.

Sources from the last 72 hours

1. [1]Why Did He Dodge the Tesla-SpaceX Merger Question?Sep 27, 2026, 5:15 AM UTC
2. [2]モルガン・スタンレーのジョナス氏：物理AIは世界GDPを倍増させる可能性 (Morgan Stanley’s Jonas: Physical AI Could Multiply Global GDP)Sep 26, 2026, 12:00 AM UTC
3. [3]The Actuator Shortage That Could Slow the Robot RevolutionSep 25, 2026, 12:00 AM UTC

AI-generated article based on recent web research, then preserved as a dated editorial snapshot.