A surge in the popularity of generative AI applications is prompting tech titans like Microsoft and Google to seek out nuclear energy to fuel their power-hungry data centers. These facilities are essential for training and operating the colossal AI models that underpin today's cutting-edge AI tools.

Both Microsoft and Google are inking agreements with U.S. nuclear energy providers to enhance the energy supply for their data centers. Recently, Google announced an agreement with Kairos Power, a company specializing in small modular reactors, to bolster the progress of AI. Michael Terrell, Google's Senior Director for Energy and Climate, stated, "The grid requires clean, dependable energy sources capable of supporting the expansion of these technologies. We believe nuclear energy can significantly contribute to meeting our energy needs in a clean, round-the-clock manner."

Google anticipates that its first nuclear reactor from Kairos Power will be operational by 2030, with additional reactors coming online through 2035. Microsoft, not to be left behind, last month entered into a deal with Constellation to reactivate a dormant reactor at the Three Mile Island nuclear power plant in Pennsylvania, which has been inactive for the past five years. The Three Mile Island facility was the site of the most severe nuclear accident in U.S. history in 1979, when a malfunctioning valve led to a reactor overheating and a subsequent meltdown.

Amazon also joined the nuclear energy bandwagon, announcing a substantial investment with Dominion Energy to explore the development of a small modular nuclear reactor near the utility's North Anna nuclear power station, in a deal worth $500 million.

The tech industry is feeling the pressure to secure energy sources for their data centers, which are a critical component of modern cloud computing and AI applications. Many developers lease servers equipped with GPUs from cloud "hyperscalers" like Amazon, Microsoft, and Google, which have seen a significant increase in demand for generative AI applications such as OpenAI's ChatGPT. This surge in popularity has led to a substantial increase in energy consumption.

According to a report by the International Energy Agency, global electricity consumption from data centers, AI, and the cryptocurrency sector is projected to double from an estimated 460 terawatt-hours in 2022 to over 1,000 terawatt-hours by 2026. A study published by researchers at the University of California, Riverside, last year found that ChatGPT consumes 500 milliliters of water for every 10 to 50 prompts, depending on the AI model's deployment location and time. As of August, OpenAI reported over 200 million users submitting questions to its chatbot ChatGPT weekly, a figure that has doubled from the 100 million active users reported in November of the previous year.

The quest for nuclear energy is not without its detractors. Many environmental activists and climate advocates oppose nuclear power due to its potential environmental and safety hazards, as well as its non-renewable nature. Greenpeace, a prominent climate charity, argues on its website that nuclear power is "incredibly expensive, hazardous, and slow to build." They contend that while it does not emit carbon dioxide or other greenhouse gases during electricity generation, it is not a viable alternative to renewable energy sources.

On the other hand, proponents of nuclear energy argue that it provides a nearly carbon-free electricity source and is more reliable than renewable energy sources like solar and wind. Rosanne Kincaid-Smith, COO of Northern Data Group, a global data center provider, expressed her views at a tech conference in London, saying, "If nuclear energy is developed and secured properly, I believe it is the future. People are fearful of nuclear due to past disasters, but I don't see traditional power grids being the sustainable energy source needed for the ongoing development of AI."

While Northern Data Group is not currently utilizing nuclear energy and is not actively exploring it as a power source for its AI data centers, Kincaid-Smith emphasized the importance of contributing to the conversation surrounding nuclear energy. "It's crucial for the broader ecosystem and the economy," she stated.

The debate over nuclear energy's role in powering the future of AI and data centers is far from settled. As tech companies continue to push the boundaries of AI and cloud computing, the search for sustainable and reliable energy sources remains a critical challenge. The potential of nuclear energy to meet these demands is a topic that will undoubtedly continue to spark intense discussion and debate in the years to come.