The rapid development of AI raises concerns and opportunities regarding global net zero goals, impacting energy demands, especially within the tech sector. Tech companies are increasingly investing in nuclear power plants to supply energy for AI data centers.
Nuclear energy as a solution for AI’s growing energy demands
Nuclear fission, a technique used in these plants, is viewed as cleaner and more reliable than fossil fuels, wind, or solar energy.
In parallel, Silicon Valley investors are channeling funds into nuclear fusion technologies. This nascent power generation method promises to produce even greater energy while yielding fewer greenhouse gas emissions and less radioactive waste. Industry experts suggest that nuclear energy may be the cornerstone to fulfill the substantial energy requirements of the AI revolution. Franklin Servan-Schreiber, CEO of nuclear energy startup Transmutex, stated, “AI requires massive, industrial-scale amounts of energy. Only nuclear power will be able to supply this massive energy demand in a reliable manner.”
However, establishing a dependable network of nuclear power plants remains a long-term ambition that will require significant financial backing and government assistance. As of August 2023, the U.S. operated only 54 nuclear power plants, as recorded by the U.S. Energy Information Administration. Major companies like Amazon and Google have entered agreements with developers of smaller, modular reactors, which can be deployed more swiftly than traditional models. Despite these efforts, physicist Edwin Lyman, director of nuclear power safety at the Union of Concerned Scientists, described current investments as merely “a drop in the bucket” compared to the many billions required.
With the urgency of AI’s energy needs, tech giants might temporarily resort to fossil fuels. Toby Rice, CEO of natural gas producer EQT, emphasized in an interview with The Wall Street Journal the limited time frame for constructing new energy infrastructure: “Tech is not going to wait 7 to 10 years to get this infrastructure built. That leaves you with natural gas.” At a recent energy conference, Rice noted that he faced repeated inquiries about the speed of operations and the availability of natural gas.
The focus on energy use at the recent UN COP29 climate summit raised eyebrows. Notably, many Big Tech firms opted not to participate prominently in the conference’s “green zone.” This decision has led some to speculate that the rapid energy consumption associated with AI data centers might jeopardize these companies’ commitments to clean energy. Kevin Thompson, COO of Gesi, a digital sustainability advocacy group, pointed out the potential pitfalls: “If our industry starts getting treated similar to oil and gas, the public relations to counter that are going to be very expensive.”
Data centers, currently powered by a blend of natural gas, coal, and renewable energy, are anticipated to increase their share of U.S. power demands from 3% to 4% to between 11% and 20% by 2030, based on a report from McKinsey. However, proponents of AI remain optimistic, believing that the ongoing intelligence revolution could instigate a corresponding energy revolution. “My hopes and dreams is that, in the end, what we all see is that using energy for intelligence is the best use of energy we can imagine,” asserted Jensen Huang, CEO of Nvidia, during an interview at Hong Kong University of Science and Technology.
The AI boom is reshaping the energy game, and nuclear power is stepping into the spotlight. With AI systems demanding unprecedented amounts of energy, nuclear fission—the process of splitting atomic nuclei—is emerging as a go-to solution for tech companies ramping up their data centers. It’s cleaner than fossil fuels and, for now, more practical than renewables like wind and solar for consistent output. But the road to integrating nuclear power is anything but smooth.
Is AI green, how sustainable is it?
While fission dominates current investments, nuclear fusion is grabbing headlines as the next big thing. By merging atomic nuclei, fusion promises even greater energy output with fewer emissions and less waste. The catch? It’s still far from ready, requiring massive funding and regulatory backing to make it viable. Even fission, despite its head start, faces challenges with expensive infrastructure, lengthy construction timelines, and lingering public skepticism.
For now, many tech giants are hedging their bets. Deals with nuclear developers for modular reactors are in motion, but the lack of immediate solutions has them turning to fossil fuels—especially natural gas—to keep their operations running. This short-term pivot raises eyebrows, as it clashes with their public commitment to sustainability and complicates their long-term green energy goals.
The stakes are especially high in public forums like the UN COP29 summit, where Big Tech’s energy use is under intense scrutiny. Skipping the usual green zone fanfare, these companies are quietly recalibrating their strategies, aware that any misstep could tarnish their environmental image. The energy demands of AI are forcing a reckoning: can tech deliver on its clean energy promises while fueling its own growth? Or will the push for innovation outpace the planet’s ability to adapt?
Featured image credit: NOAA/Unsplash
