Microsoft is taking a risk by exploring the world of nuclear power and using artificial intelligence (AI) to simplify the difficult process of putting up nuclear reactors. This move is intended to fuel the company’s energy-hungry AI models. The IT behemoth is looking for sustainable alternatives to conventional fossil fuels like coal and natural gas because it recognizes the enormous energy requirements of machine intelligence.
The Wall Street Journal reveals that Microsoft has dedicated the past six months to training a large language model with U.S. nuclear regulatory and licensing documents. The goal is to expedite the extensive paperwork required for approvals to establish new nuclear power plants—a process that can span years and incur costs in the hundreds of millions of dollars. This audacious endeavor underscores Microsoft’s commitment to AI, as the technology is now being entrusted to navigate the intricacies of regulatory compliance.
Rather of depending solely on renewable energy sources like wind and solar power, Microsoft and other major IT companies like Google are focusing on small modular reactors (SMRs). By standardizing construction, these smaller power stations hope to lower construction costs and increase the economic viability of nuclear energy. There are currently no operating SMRs in the United States, despite the fact that many global designs are being considered. This is mostly because of the prohibitively high costs related to permits and construction.
To tackle these challenges, Microsoft has partnered with the nonprofit Terra Praxis. According to Terra Praxis co-CEO Eric Ingersoll, AI could slash up to 90 percent of the human hours traditionally required to secure approval for a new nuclear plant. Addressing concerns about potential AI mishaps in the delicate field of nuclear regulation, Ingersoll reassures that the AI is meticulously trained on highly structured documents, minimizing the risk of erratic behavior.
As Microsoft pioneers the fusion of AI and nuclear power, it remains at the forefront of technological innovation, aiming not only to meet its insatiable energy demands but also to revolutionize the way we approach the complexities of nuclear energy regulation.