The Biden administration recently announced plans to work with tech firms on power usage. Technological revolutions have always come with novel challenges: building infrastructure, navigating dual-purpose applications and preparing our workforce. For the growing industry surrounding AI, this challenge is energy. Progress in AI will soon face the limits of the power available on the grid. As the administration now recognizes, maximizing AI’s potential will require an energy breakthrough. While the nation needs diverse approaches, the best solution is the promise of fusion.
Currently, AI models are run from data centers. These models require power to build, or “train,” and to run, or “inference.” Both are energy-intensive. Combined, the vast expansion of AI means that the electricity consumption of building and running AI models may soon match the power generation of small countries — and then grow beyond that. Estimates project that a single utility in Virginia will see annual data center power demand grow by nearly 79,000 GWh by 2038 — the equivalent of adding nearly 6.1 million homes to the grid.
To meet that demand, AI firms and utilities will need to turn to new technologies. Fusion energy, the process that powers the sun and stars, holds the potential to provide a clean, safe and virtually limitless source of energy. Unlike fossil fuels, it doesn’t release greenhouse gases, and unlike nuclear fission, it doesn’t generate dangerous radioactive waste, nor does it bring a similar risk of accidents.
Skeptics contend that fusion has long promised much but failed to deliver. Historically, that has been true. But there is reason to believe this moment is different. Recent advancements in this field are promising. Lawrence Livermore National Lab has repeatedly broken power generation records, and achieved net energy, an important milestone.
Around the globe, investment in fusion is accelerating. China spends an estimated $1.5 billion per year on fusion, compared to the U.S. government’s $763 million in 2023. The U.K., Germany and Japan have all announced new investments and strategies. The private sector has invested over $6 billion in fusion start-ups.
Success in AI itself may also enable fusion science. In February, a team from Princeton announced that they had utilized a specialized AI model to significantly improve the stability of the superheated plasma in a fusion reactor — an important requirement for a successful and commercially sustainable fusion reactor.
For America, fusion success is also a national security issue. Our adversaries weaponize energy dependence, using it as a tool for coercion. A clean, domestic source of energy like fusion would not only free us from this geopolitical stranglehold but also empower our allies. Take Taiwan, for instance, a global leader in hardware manufacturing. By partnering with Taiwan, we could not only leverage their expertise in building the complex machinery needed for fusion reactors but also address their vulnerability to energy blockades imposed by China.
To turn recent fusion breakthroughs into near-term success, the United States will need a moonshot mentality, with sustained research funding and mission-focussed leaders. Building the infrastructure for large-scale fusion reactors is no small feat. We will need to resolve complex engineering and scientific challenges, building reliable reactors that are cost-effective and commercially viable.
A moonshot approach would mean learning from our greatest scientific successes in the past. Without a moonshot, humanity may have eventually gone to the moon or harnessed the power of the atom. But the organized focus of the American government meant that these goals happened in a matter of years, not decades — and that their benefits were redounded to us.
In the case of fusion, the challenges pale in comparison to the potential rewards. A successful fusion moonshot would revolutionize our energy landscape, paving the way for a sustainable future. It could free us from the shackles of fossil fuels, mitigating climate change and its devastating consequences. It would create new industries and jobs, driving economic growth and innovation. Most importantly, it would be a key ingredient in the development of increasingly effective AI models — allowing them to deliver their promise of prosperity and catalyze a flywheel of innovation across many technologies.
As the constellation of emerging technologies — AI, robotics, biotechnology — progresses we will need a holistic approach to technological development. A successful fusion moonshot wouldn’t exist in a vacuum; it would act as a catalyst for further technological breakthroughs, fostering collaboration and accelerating progress across the board.
The history of technological revolutions teaches us that challenges are inevitable. Yet it is through embracing these challenges and daring to dream big that we usher in eras of unprecedented progress. AI promises to change the world — a fusion moonshot would power that future. A moonshot would harness the American spirit of innovation, scientific breakthroughs, commercial investment and geopolitical imperative — and turn the revolutionary promise of fusion into reality.
