Build Rate, Not Energy Reserves, Will Determine AI Leadership
Philip Zelikow of the Hoover Institution argues that U.S. leadership in AI will depend not only on chips and software but on the ability to build electricity generation and transmission at far greater speed. America’s fuel resources and nuclear prospects offer little strategic advantage, he says, if projects take 10 to 15 years while China expands power capacity across multiple sources at scale. Zelikow calls for bipartisan regulatory changes that turn energy abundance into operating capacity for a data-center-intensive economy.
The advantage is not energy in the ground but power built in time
Philip Zelikow argues that energy is becoming a strategic constraint on American leadership in the AI phase of the digital revolution. The United States has abundant fuels and promising technologies, including fourth-generation nuclear systems and small modular reactors. But those assets confer no decisive advantage, in his account, unless the country can turn them into operating generation and transmission fast enough.
China is the counterpoint. Zelikow says China has already installed electricity-generation capacity across coal, hydro, renewables, and nuclear at unparalleled scale. It also has the potential to supply Chinese data centers with domestic AI accelerator chips. That combination makes build rate—not just resource endowment—the relevant measure of readiness.
We can't tolerate 10 or 15 years for a nuclear power plant or a power transmission line. That just won't be good enough.
His prescription is not simply to produce more energy. It is to “build better”: to make approval and construction timelines compatible with the scale and speed of the technological competition he describes. Without that, American energy abundance remains potential rather than usable capacity for a data-center-intensive economy.
AI leadership rests on hardware, software, and power
Zelikow places energy alongside hardware and software as the three critical enablers of success in the digital revolution. Artificial intelligence, he says, is that revolution’s culminating phase—a transformation that began significantly in the 1980s and accelerated substantially in the 2000s.
The hardware component is compute chips. American firms are central to their design, but Zelikow emphasizes that production depends on an international ecosystem: partners in Japan and the Netherlands, and above all the ecosystem centered in Taiwan. A Semiconductor Industry Association map shown with his remarks locates U.S. semiconductor research, design, manufacturing, equipment, materials, and university partners across the country.
Software follows a similar pattern. Much software design and engineering takes place in the United States, Zelikow says, while overseas contributors remain important.
Energy is the third critical enabler for success in the digital revolution.
Energy occupies a distinct role in this system. Hardware supply and software development are internationally distributed even where U.S. capabilities are central. Zelikow’s argument is that domestic energy abundance could give the United States a comparative advantage in supplying the physical power demand of the emerging digital economy.
The industrial precedent is a system, not a single breakthrough
Zelikow looks to the second phase of the Industrial Revolution for the structure of the present challenge. He attributes American success to three major conditions: geology sufficient to supply needed materials, a society that cultivated engineering talent to apply those resources to leading technologies, and an education system that trained a workforce to operate industrial factories.
He adds an enabling political environment that pulled those capacities together. In his telling, that combination helped the United States win the Second World War and become a dominant superpower.
The analogy is not that AI repeats industrialization. It is that leadership in a major economic transformation cannot be reduced to a technical invention. Resources need engineers able to apply them, a workforce able to operate the resulting technologies, and political arrangements that allow the whole system to be built.
That framing puts energy in context. Hardware, software, and power are not separate checkboxes in Zelikow’s argument. Their value depends on whether the country can organize and scale them as part of a functioning industrial system.
The political opening matters only if it changes the rules for building
Zelikow sees a potentially unusual overlap between the two major U.S. parties. He says that, for the first time in half a century, both are aligned around removing obstacles that have blocked approvals and constrained the country’s ability to build.
He describes the left as having an “abundance agenda” and the right as seeking to build in America while freeing private enterprise from government overreach. On screen, a The Hill opinion piece calls to “unleash the ‘Abundance Agenda,’” while a Forbes headline refers to growing bipartisan support for protecting the freedom to invest.
But his desired outcome is more specific than shared pro-building sentiment. The task, he says, is to devise particular political changes that can command bipartisan support and revise a regulatory environment that has kept the country from building. Those changes must affect the approvals and constraints that determine whether generation and transmission projects take years or more than a decade.
He frames the opportunity in broader strategic terms: Americans recognize a period of global uncertainty and share a national goal of winning the AI revolution. Energy abundance at home, combined with partnerships abroad, could position the United States centrally in the emerging economic age. The political test is whether that alignment produces rules capable of unlocking the energy potential he says already exists.