The AI Trust Presents Powering Progress – Executive Summary

On July 31, the AI Trust Foundation kicked off its energy & AI initiative at the U.S. Capitol with “Powering the Future: An Outlook on Energy and Artificial Intelligence.” Moderated by the Hon. Jerry McNerney, Chairman of the AI Trust Foundation, the event featured leaders from Groq, Lawrence Livermore National Labs, NEI, EEI, and The AI Trust Foundation on the pivotal role of AI in revolutionizing the energy sector. Event participants included executives from ITC Holdings, TypeOne Energy, Meta, Box, Inc., NationalGrid, Hitachi, and Pillsbury Winthrop.

Key Insights From the Event

• AI Driven Energy Solutions: Aileen Black, President of Groq, highlighted the need for fast and efficient AI models to predict and adapt to energy demand fluctuations, ensuring a sustainable energy supply.
• Infrastructure Investment: Phil Dion, Senior Vice President of Customer Relations at Edison Electric Institute, emphasized the critical need for increased electric capacity and reliable infrastructure to support AI growth.
• Nuclear Energy’s Role: Benton Arnett of the Nuclear Energy Institute discussed the importance of nuclear power in achieving carbon-free energy goals.
• Collaborative Partnerships: Dr. Colin Ponce of Lawrence Livermore National Laboratory stressed the importance of partnerships between government, industry, and academia to address AI and energy challenges.

Panel Recommendations:

• Develop and implement AI tools to optimize energy consumption and distribution.
• Advance regulatory reforms to de-risk investment in support of the development and implementation of energy solutions to meet the growing AI energy demand.
• Foster partnerships between government, industry, and research to drive innovation and policy development.
• Invest in nuclear power infrastructure and research to achieve carbon-free energy goals.
• Prioritize talent development, education, and high-skilled immigration in AI and power systems to meet industry demands.

AI-driven Energy Revolution

Aileen Black, President of Groq – an American manufacturer of highly efficient Language Processing Units (“LPU”) for computationally intense AI applications, emphasized growing energy consumption of AI and how it can be managed. AI models can help predict and adapt to fluctuations in energy demand, ensuring a reliable and sustainable supply. However, she warned that because of the AI-fueled demand in energy growth, there is a need for investment in infrastructure and the development of new technologies to support AI’s widespread adoption. Tokens, the metric used in AI models to compute demand, are the new tool that indicates energy usage. The new processor technology, designed specifically for AI, is ten times more efficient than current technology, and helps unlock the black box of understanding how AI develops its output.

Phil Dion, Senior Vice President of Customer Relations at Edison Electric Institute, highlighted the increasing AI demand for electric energy, and the need to reducing the risk for private investment in electric infrastructure. Claiming that the electric grid was the greatest engineering achievement of the 20th Century and the need for it to maintain that distinction in the 21st Century, he suggested that AI can help optimize energy distribution and enable the development of decentralized power systems. The Information Age needs power and scale never before seen and AI can provide the accurate forecasting needed for infrastructure investment. Dion underscored the importance of partnering with customers in infrastructure plans. He spoke of the need to carefully consider the tradeoffs between reliability and marginal reductions in emissions in a period of explosive growth in energy demand. He also indicated that the supply chain challenges will play heavily in future reliability.

Benton Arnett of the Nuclear Energy Institute, spoke about the value of nuclear power in achieving carbon-free energy. He emphasized the technology’s reliability, dispatchable power, and consistent reliability, making it an attractive option for data centers and corporate sustainability goals at a broad range of scale. Arnett also discussed the challenges facing nuclear power, including regulatory hurdles and public perception. He noted that the nuclear industry has not experienced a surge in buildouts since the 1970s and that new, first-of-a-kind reactors will require dedicated capital for risk management that should not be borne by utilities alone. Arnett also emphasized the distinct roles that new nuclear technologies will play in an evolving energy demand landscape. For example, high temperature gas reactors could enable advanced forms chemical processing and natural gas reforming, while 50 to 300 MW small modular reactors (“SMRs”) and up to 50 MW microreactors could add distributed generation to the power grid while enabling the oil and gas industry to reduce its reliance on diesel in the Texas Permian Basian, the Athabasca oil sands, and elsewhere. He forecast the need for 200 GW of more of nuclear power in the next decades.

Dr. Colin Ponce of the Lawrence Livermore National Laboratory explained how the challenges of AI are not business as usual with explosive demand growth, heightened security issues, and climate change. AI can benefit the power grid by ensuring safety, reducing waste, and improving reliability. He noted that the need for a robust AI partnership between government, industry, and academia is crucial for meeting the 21st-century goals of the power grid. He emphasized that policies are needed to mitigate risk.

All speakers agreed that policy based on an understanding of the technology and its consequences that provides predictability is needed for the United States to maintain its innovation leadership.

The panel also touched on the crucial issue of talent and education in the AI and energy sectors. Dr. Ponce emphasized the need for a diverse workforce, with a strong emphasis on developing skills in both AI and power systems. Aileen Black highlighted the importance of encouraging young people to pursue careers in power systems, noting the exciting potential of jobs in AI.

July 31st, U.S. Capitol, “Powering Progress: Energy & AI” featuring Benton Arnett, Nuclear Energy Institute (NEI); the Hon. Jerry McNerney, Chairman of the AI Trust Foundation; Aileen Black, President, Groq; Philip Dion, Senior Vice President, Edison Electric Institute (EEI); Dr. Colin Ponce, Lawrence Livermore National Labs.