Elon Musk Has Given Up on Solar Power (on Earth)

Background and Context

By May 2026, a fundamental reversal in energy strategy has emerged as the most significant development within Elon Musk’s corporate ecosystem, marking a decisive departure from his historical identity as a proponent of clean energy. For years, Musk’s business ventures, including the founding of Tesla and the acquisition of SolarCity, were explicitly designed to accelerate the global transition to sustainable energy sources. However, recent operational shifts indicate that his artificial intelligence subsidiary, xAI, is now aggressively deploying natural gas combined-cycle power facilities to support its rapidly expanding computational infrastructure.

Simultaneously, SpaceX has redirected its focus away from terrestrial photovoltaic arrays, accelerating the prototype validation of orbital data centers instead. This pivot signals a transformation in Musk’s approach from an idealistic advocate of a solar-driven economy to a pragmatic realist prioritizing computational capacity above all else. The timeline reveals that since the expansion of xAI’s Groq cluster began in the second half of 2024, the demand for power stability has exceeded the capabilities of solar and battery storage combinations, forcing leadership to reevaluate their energy mix. This is not an isolated incident but rather a microcosm of the broader AI infrastructure industry’s collective response to energy bottlenecks, highlighting the tension between technological acceleration and physical resource constraints.

Deep Analysis

The core logic behind this strategic shift lies in the structural mismatch between AI computational density and the characteristics of energy supply. Modern large language model training and inference require electricity that is high-load, highly stable, and immediately available. Solar power generation suffers from inherent intermittency and volatility. Even when paired with the most advanced lithium-ion battery storage systems, the cost-effectiveness and reliability of solar face significant challenges when dealing with continuous peak loads in the megawatt or gigawatt range. In contrast, while natural gas generation has higher carbon emissions, it offers advantages such as rapid start-stop capabilities, stable output, and relatively controllable unit energy costs, making it an ideal baseload or peaking power source. For xAI, ensuring that GPU clusters operate continuously around the clock is fundamental to business survival; any training interruption caused by grid fluctuations could result in losses amounting to millions of dollars. Therefore, the choice of natural gas is not a betrayal of environmental principles but a rational calculation based on return on investment and operational continuity.

Furthermore, SpaceX’s vision for orbital data centers represents a solution in a different dimension: leveraging the nearly infinite solar resources in space, combined with wireless energy transmission technology, to solve the problems of energy transmission loss and land occupation on Earth. The雏形 of this "space energy internet," although technically extremely difficult, may be more disruptive in the long term than distributed photovoltaics on the Earth's surface. This approach attempts to decouple energy generation from land constraints, offering a potential pathway to sustain the exponential growth of AI workloads without the geographical and environmental limitations of terrestrial renewable energy projects. The technical complexity involves not only generating power in orbit but also transmitting it efficiently to ground stations or other satellites, a challenge that requires breakthroughs in both materials science and wireless power transfer technologies.

Industry Impact

This strategic adjustment has profound implications for the competitive landscape of the industry and its user base. First, it intensifies the "energy arms race" in the AI infrastructure sector. As major players like xAI, OpenAI, and Google DeepMind lock in large natural gas power plants or nuclear projects, energy supply has become a critical bottleneck constraining AI development, rather than just IT hardware. Traditional energy giants such as ExxonMobil and Shell have thus gained new strategic opportunities by providing customized energy solutions for tech giants, reshaping their position in the digital economy. For cloud service providers, this means that the definition of "green computing power" may need to be redefined, and the short-term increase in carbon footprint may become an "dirty transition" that the industry has to accept. The shift also creates a cognitive shock for ordinary consumers and environmental advocates. Musk’s personal brand光环 of environmental protection has faded, and the public may question the double standards of his business empire regarding climate change responsibility.

However, from a market perspective, this pragmatic strategy ensures the leading position of its enterprises in the AI wave, proving that commercial feasibility often takes precedence over ideological purity in the process of technology implementation. In terms of competitive landscape, AI startups that have failed to secure stable energy supplies in advance will face greater survival pressure, and the industry concentration may further increase. The reliance on natural gas and nuclear power introduces new dependencies on fossil fuel markets and regulatory environments, potentially exposing these tech giants to volatility in energy prices and stricter environmental regulations. This shift also forces a reevaluation of the environmental impact of AI, as the carbon intensity of training large models becomes a more prominent issue in public discourse and corporate sustainability reporting.

Outlook

Looking ahead, Musk’s energy strategy shift will trigger a series of subsequent developments worth watching. In the short term, the implementation speed of xAI’s natural gas projects and their actual carbon emission data will become the focus of industry observation, providing a reference benchmark for other AI companies. In the medium term, with breakthroughs in solid-state battery technology and the commercialization of small modular reactors (SMRs), the energy structure of AI data centers may undergo dynamic changes again, with natural gas potentially existing only as a transitional solution. In the long run, if SpaceX’s orbital data centers achieve technical breakthroughs, they will completely change the geographical distribution logic of global energy and computing power, achieving the decoupling of energy and computing power.

It is worth noting that Musk has not completely given up on solar power, but has shifted his focus from "ground distributed generation" to a combination of "space centralized generation" and "high-efficiency energy storage technology." This evolution reminds us that the energy transition is not a linear process, but a result of continuous博弈 between technical feasibility, economic costs, and political will. For investors and industry observers, understanding the deep logic behind this "pragmatic shift" is more important than simply judging its moral stance, as it heralds a new paradigm of technology and energy integration in the next decade. The success of these initiatives will likely determine not only the future of Musk’s companies but also the trajectory of the global AI industry’s energy consumption patterns.