Norman and the NERD: A Google Insider Reveals How the Search for Clean Nuclear Energy Went Awry at One of the World's Largest Tech Companies, with a Strange Detour into Cold Fusion.
The NERD program transformed nuclear energy policy and promoted innovation in this field.
Google has been involved in a series of innovative developments, and its foray into nuclear energy stands out as a lesser-known chapter in its history. This exploration, which included an unexpected diversion into cold fusion and machine learning, was conducted through partnerships, such as the one established with TAE to enhance Norman, a machine intended for advanced plasma experiments. Google’s NERD (Nuclear Energy R&D) program focused on researching clean nuclear energy, facilitating a series of initiatives that have garnered attention in the tech field.
Google's interest in nuclear energy was not limited to the technical realm; it also sought to influence policies related to the sector. In collaboration with think tanks like Third Way, the company worked to establish a regulatory environment that favors the development of advanced nuclear reactors. These designs, which promise greater safety, efficiency, and better waste management methods, are often hindered by outdated regulations. To counter this challenge, the NERD team supported legislative actions that resulted in laws aimed at modernizing nuclear licensing processes and providing funding for demonstration projects.
The initiative proved successful, as the U.S. federal government committed billions to support a new generation of nuclear reactors, including small modular reactors (SMRs). Recently, Google signed an agreement to procure nuclear energy from SMRs, highlighting its practical commitment to nuclear revival.
In the field of nuclear fusion, Google combined its computational capability with plasma research at TAE Technologies. A key element of this collaboration was Norman, a reactor designed to heat plasma to extreme temperatures. A machine learning tool known as the "optometrist algorithm" played a crucial role in optimizing the reactor's settings, enabling physicists to refine experiments efficiently, achieving significant advances in plasma stability and temperature control.
These efforts led TAE to reach new milestones, as Norman exceeded its initial goals. The lessons learned in this process spurred the development of Copernicus, a reactor aimed at achieving "energy gain," where energy production equals energy consumption.
However, one of NERD's more unusual projects was the investigation of low-energy nuclear reactions (LENR), commonly known as cold fusion. Although the credibility of cold fusion was undermined by controversial claims in the 1980s, Google approached the topic with a rigorous scientific focus. By funding 12 projects, the company sought to verify if anomalies in previous experiments could lead to significant advancements. The result was a lack of evidence for cold fusion, but an abundance of peer-reviewed publications and unexpected applications.
Google's ambitions range from transforming nuclear policy in the U.S. to employing machine learning in fusion, also reexamining discredited concepts. This unconventional journey reflects the company's broader philosophy: no idea is too big, bold, or controversial to be explored. Although not all initiatives achieved their objectives, each contributed to a greater understanding of the potential of nuclear energy.
