The Department of Energy (DOE) has released a new roadmap for the United States to realize its decades-long dream of harnessing fusion energy.
This is a commitment to supporting research and development efforts and pursuing public-private partnerships to finally build the first generation of fusion power plants. And of course, the plan presents AI as both a tool that can lead to new advances and as a motivation to create a new energy source that can satisfy the growing electricity demand of data centers.
The DOE is considering an extremely ambitious timetable, although the details of how to get there are vague, given that success still relies on achieving scientific breakthroughs that have eluded scientists for the better part of a century. What’s more, the burgeoning ecosystem of startups and researchers engaged in this task is crying out for more money — funds that the DOE admits it doesn’t yet have to give.
Of course, the plan highlights AI
A DOE press release yesterday boasted that its new strategy aims to deploy fusion energy commercially on power grids by the mid-2030s. The current roadmap, however, paints a fuzzier picture. The document states in bold that its goal “is to provide the public infrastructure that supports the merger of the private sector in the 2030s.” Regardless, there are still many hurdles and uncertainties to overcome, which could realistically lead us to powering our homes and businesses with fusion energy decades from now, if not a day.
Why is this such an important task? Today’s nuclear fission plants split atoms to release energy. Nuclear fusion plants, on the other hand, would fuse atoms to generate energy in a controlled manner. (You get a hydrogen bomb when this is done uncontrolled.) The advantage of doing fusion would be that it doesn’t produce the same radioactive waste as fission, nor does the process rely on polluting fossil fuels.
Fusion essentially mimics the way stars produce their own light and heat. While this can provide an abundant carbon-free energy source, it also requires an enormous amount of heat and pressure to fuse the atoms. As a result, it has been extremely difficult to achieve a fusion reaction resulting in a net energy gain (what is called “ignition” in industrial parlance). Scientists achieved this for the first time in 2022 using lasers. Researchers developing fusion technologies are working to recreate this feat and figure out how to keep the reaction going longer.
There have been other significant changes in recent years that have fueled all of the current buzz around the merger. The generative AI boom has big tech companies scrambling to get enough electricity to power more data centers. Sam Altman, Bill Gates, and Jeff Bezos have all backed fusion startups developing their own factory designs. Google and Microsoft have announced plans to purchase electricity from future fusion plants expected to be operational by the late 2020s or 2030s. More than $9 billion in private investment has gone into fusion demonstrations and reactor prototypes, according to the DOE.
There are other big gaps to fill, and that’s where the DOE says it can step in. The roadmap emphasizes bringing the public and private sectors together to build the “critical infrastructure” needed to make fusion commercially viable, such as the production and recycling of fusion fuels (usually isotopes of hydrogen called tritium and deuterium). Another “key challenge” highlighted in the paper is the need to develop structural materials strong enough to withstand the extreme conditions of a smelter plant. (Remember that you are somehow replicating the environment within a star.)
He also mentions the development of regional hubs for fusion innovation, where DOE labs could work with universities, local and state governments, and private companies to build a workforce for these new technologies. One hub would be a collaboration between Nvidia, IBM, the Princeton Plasma Physics Laboratory and the DOE to “establish an AI-optimized fusion-centric supercomputing cluster” called Stellar-AI.
The DOE devotes an entire section of the roadmap to AI, which it calls a “transformative tool for fusion energy.” Researchers can use AI models to build “digital twins” to more quickly study the performance of experimental facilities, the roadmap says as an example.
The document also comes with a big warning. Written at the top, above the summary, it states: “This roadmap does not commit the Department of Energy to specific funding levels, and future funding will be subject to Congressional appropriations.” » In other words, the DOE is not yet ready to invest money in this project.
And while the Trump administration has incorporated fossil fuels, nuclear fission and fusion into its ambitions for “energy dominance,” the president has clawed back funding for solar and wind power projects that are already much faster and generally cheaper to deploy to meet America’s growing electricity demand.
