The U.S. is facing rising energy demand driven by artificial intelligence and global competition. Many experts argue that current infrastructure cannot keep pace without major changes.
Kevin Kong, founder and CEO of Everstar, says nuclear power is the only scalable solution to meet that demand while maintaining reliability and affordability.
Kong frames energy development as the foundation of human progress. “Our history as a species has just been one long history of unlocking more and more energy,” he says. He argues that nuclear power represents the next leap, calling it “the only one that’s reliable, dense, clean enough to power high advancements for the next few decades.”
Safety concerns remain a barrier, though Kong says modern reactors are fundamentally different from earlier designs. “There have been decades of safety engineering advancements,” he says, adding that lessons from past incidents have been incorporated into new systems. He argues that nuclear energy is “the safest source of energy” on a per-kilowatt-hour basis and highlights its efficiency compared to other sources.
Rising demand from artificial intelligence is a central concern. Kong notes that current AI usage is still in its early stages. “We are only at day one usage,” he says, pointing to the coming expansion of video, industrial applications, and advanced modeling. That growth will require significantly more power. “We expect a whole lot more token-intensive compute,” he says, warning that existing infrastructure will struggle to keep pace.
He says nuclear power is essential to meeting that demand while maintaining affordability. “We can build more natural gas… but for the long term… we really need to get our energy grid in order,” he says. He points to China’s rapid expansion, noting that it is building “six to eight reactors per year” in preparation for future energy needs.
Kong also highlights the potential of small modular reactors and microreactors. These designs allow for decentralized power generation and faster deployment. “You can put these reactors on a truck bed or build them in modules very quickly,” he says. He adds that smaller systems could serve data centers, remote sites, and defense applications.
Regulation remains an obstacle. Kong says the current permitting process is slow and costly, citing one reactor approval that “took half $1 billion and spread across ten years of effort” and required “on the order of 2 million documents.” He contrasts nuclear with other industries, noting that “you cannot do anything until the approvals come out.” That dynamic, he says, discourages private investment and delays innovation.
According to him, regulatory growth since the 1979 Three Mile Island incident has driven costs higher. “It’s exploded out of control,” he says, adding that companies face “a decade of project overruns and billions of dollars in cost overruns.” He believes reform is necessary to make nuclear development economically viable while maintaining safety standards.
Kong also emphasizes the geopolitical stakes. He warns that China’s progress in reactor design and deployment could reshape global energy markets. “Each reactor represents a 100-year diplomatic friendship,” he says, noting that countries adopting foreign nuclear technology often enter long-term partnerships. If the United States fails to compete, he says, “it’s going to be obvious who they have to go to, and it’s not going to be America.”
Younger generations, according to him, are more open to nuclear energy. “Gen Z loves nuclear,” Kong says, describing them as less influenced by past fears and more focused on data and environmental outcomes.
Kong believes the U.S. must act quickly to regain leadership. “We have built zero net new reactor since 1979,” he says, arguing that the country needs a coordinated national effort. He calls for government support to reduce risk and accelerate construction, comparing the challenge to past industrial mobilizations. “America has done incredible things before,” he says. “We put people on the moon.”
Everstar aims to address part of the problem by using AI to streamline regulatory processes and improve efficiency. Kong says the company has already demonstrated “90% timeline compression” in some applications. He sees that as one step toward a broader transformation of the industry.
