⚡NIMITZ TECH NEWS FLASH⚡

“America’s AI Moonshot: The Economics of AI, Data Centers, and Power Consumption”

House Oversight Subcommittee on Economic Growth, Energy Policy, and Regulatory Affairs

April 1, 2025 (recording linked here)

HEARING INFORMATION

Witnesses and Written Testimony:

• Mr. Neil Chilson: Head of AI Policy, Abundance Institute

• Mr. Josh Levi: President, Data Center Coalition

• Mr. Mark P. Mills: Executive Director, National Center for Energy Analytics

• Mr. Tyson Slocum: (Minority Witness): Energy Program Director, Public Citizen

HEARING HIGHLIGHTS

IN THEIR WORDS

SUMMARY OF OPENING STATEMENTS FROM THE COMMITTEE AND SUBCOMMITTEE

• Chairman Burlison highlighted AI’s broad benefits in healthcare, manufacturing, and defense, and praised the Trump administration’s $500 billion Stargate initiative and Executive Order 14179 to secure U.S. AI leadership. He raised concerns about the rising energy demands of data centers, projecting usage could triple by 2028. He stressed the need for skilled workers and urged Congress to support policies that enable America to lead the global AI race.

• Ranking Member Frost supported continued AI investment but criticized the removal of civil rights safeguards and the prioritization of corporate interests. He warned of environmental and social harms, including tax breaks with few jobs, increased emissions, and AI misuse in the creative sector. Frost raised concerns about biased AI in healthcare and national security risks like deep fakes. He called for balanced policy that fosters innovation while protecting people, equity, and the environment.

SUMMARY OF WITNESS STATEMENT

• Mr. Chilson argued that AI is already delivering real benefits in healthcare, productivity, and entrepreneurship and could generate $20 trillion in global economic impact by 2030. He urged Congress to preserve a light-touch, sector-specific regulatory approach, citing Utah's AI Act as a strong model. He warned against a patchwork of conflicting state laws and called for energy regulatory reform to accelerate grid expansion and support AI infrastructure.

• Mr. Levy emphasized the central role of data centers in powering AI and the digital economy, noting $3.5 trillion in GDP contributions and 4.7 million jobs supported since 2017. He highlighted growing demand and industry investment across 23 states. To meet future needs, he urged Congress to streamline energy permitting, address equipment and labor shortages, and invest in workforce training.

• Mr. Mills called AI a transformative force that could boost U.S. productivity and add $10 trillion to the economy by 2030. He highlighted massive private investment in data centers and projected a need for 70–130 gigawatts of new power capacity. He argued most of this will come from natural gas and urged the government to eliminate regulatory barriers to rapid infrastructure deployment.

• Mr. Slocum warned against exaggerated projections of AI energy needs, citing past overestimates and ongoing efficiency gains. He criticized the use of emergency powers to override local control and subsidize fossil fuel infrastructure. He opposed expanded LNG exports and tariffs, arguing they raise energy costs and disrupt supply chains. Slocum called for smarter trade policy and stronger regulatory oversight.

SUMMARY OF KEY Q&A

• Chairman Burlison asked what is preventing the U.S. from reaching its full AI potential and why Americans should be more excited about AI rather than fearful. Mr. Chilson cited outdated energy permitting and conflicting state software regulations that slow innovation and deployment. He also emphasized transformative benefits of AI in healthcare and trillions in economic growth.

  Chairman Burlison also asked for examples of regulatory barriers facing data centers. Mr. Levy pointed to long permitting delays—up to seven years for transmission—and equipment shortages as key obstacles.

• Ranking Member Frost questioned why data centers consume so much energy and what AI’s impact on working families is. Mr. Slocum explained that massive numbers of processors require constant cooling and power but said demand response strategies are underused. He warned of pollution, minimal job creation, and rising utility bills from tax deals favoring big tech.

  Ranking Member Frost also asked for solutions to mitigate environmental harms. Mr. Slocum recommended clean energy sources and battery storage.

• Rep. Palmer asked about future power needs for AI, and Mr. Mills cited FERC estimates of a 60–130 GW increase by 2030 due to AI, manufacturing, and electrification.

  Rep. Palmer emphasized the threat of losing the AI race to China and suggested restarting retired coal and gas plants using small modular reactors (SMRs). Mr. Mills supported nuclear energy but cautioned that SMRs don’t yet exist at scale and would take years to deploy, requiring major infrastructure investment.

• Rep. Subramanyam did not ask any question, but shared concerns from his Virginia district, which houses the highest concentration of data centers globally, about their impact on energy bills, environmental quality, and national security. He warned that data centers are straining local resources, threatening historic sites, and making the region a strategic target. He called for a national strategy to manage data center growth thoughtfully and protect communities from unintended consequences.

• Rep. Ansari asked how data centers are improving energy efficiency. Mr. Levy explained that modern centers centralize computing to reduce waste and now tolerate higher temperatures to reduce cooling demands.

• Rep. Perry asked why Microsoft chose to power its data center with nuclear energy from Three Mile Island. Mr. Chilson said it offered a fast, reliable supply.

  Rep. Perry emphasized the importance of base-load nuclear power for AI and askedwhat the U.S. can learn from China, who builds nuclear plants quickly and still relies heavily on coal. Mr. Mills said China prioritizes cheap, fast energy to fuel its dominance in solar and AI. Both agreed that Microsoft likely chose nuclear for its affordability and reliability.

• Rep. Higgins asked whether the U.S. grid could support AI-related power consumption. Mr. Mills said no—modernization and off-grid solutions are necessary.

  Rep. Higgins also asked about water needs, and Mr. Mills said AI chips run extremely hot and require substantial cooling.

  Rep. Higgins asked about AI supply chains. Mr. Levy explained there are serious shortages in electrical components and expressed strong support for onshoring production.

• Rep. Fedorchak asked how efficiency could offset skyrocketing power demand. Mr. Mills said compute efficiency is improving rapidly, but demand is rising faster, necessitating on-site and independent power generation. Mr. Levy added that data centers are more energy efficient than decentralized computing and support broader efficiency through smart technologies like thermostats and grid monitoring.

• Rep. Boebert asked if AI and coal are vital to rural economies. Mr. Chilson said AI offers widespread economic benefits, especially for upskilling workers.

  Rep. Boebert asked if wind and solar alone can meet future power demands, and Mr. Mills said they cannot, emphasizing the need for gas turbines and a blended approach.

  Rep. Boebert warned that energy policies discouraging fossil fuels are slowing AI progress, and Mr. Mills agreed, noting that distorted subsidies have delayed needed power plant development. They concluded by discussing the vast power needs of data centers, comparing one computer rack’s energy use to 50 EVs.

“From Chalkboards to Chatbots: The Impact of AI on K-12 Education”

House Education & Workforce Subcommittee on Early Childhood, Elementary, and Secondary Education

April 1, 2025 (recording linked here)

HEARING INFORMATION

Witnesses and Written Testimony:

• Dr. Sid Dobrin: Chair, Department of English, University of Florida

• Dr. Julia Rafal-Baer: CEO, ILO Group

• Ms. Erin Mote: CEO, InnovateEDU/EDSAFE AI Alliance

• Mr. Chris Chism: Superintendent, Pearl Public School District

HEARING HIGHLIGHTS

IN THEIR WORDS

SUMMARY OF OPENING STATEMENTS FROM THE COMMITTEE AND SUBCOMMITTEE

• Chairman Kiley noted the rapid pace of AI development and its transformative potential in education. He argued that AI can address longstanding challenges, such as achievement gaps and declining national educational outcomes. Citing tools like Khan Academy’s “Conmigo,” he highlighted AI’s ability to deliver personalized, accessible tutoring. He acknowledged risks including student misuse, privacy concerns, and screen overexposure but emphasized the importance of responsible adoption. Kiley stressed that AI should not replace human interaction in learning and rejected the need for new federal mandates, advocating instead for local flexibility. He framed the hearing as an opportunity to spotlight AI’s promise and share best practices.

• Ranking Member Bonamici emphasized that AI is already impacting classrooms and must be addressed with attention to equity and regulation. She defended the Department of Education’s role in closing achievement gaps, supporting underfunded schools, and guiding safe AI integration. Bonamici criticized funding cuts to federal education programs and warned they could worsen disparities and widen the digital divide. She stressed the need for federal leadership to prevent AI from reinforcing existing biases and to support research-based decision-making. Bonamici urged bipartisan cooperation to preserve federal investments and ensure all students benefit from emerging technologies.

SUMMARY OF WITNESS STATEMENT

• Dr. Dobrin described generative AI as a turning point in education and urged its integration across subjects, not just in computer science. He argued that AI literacy builds critical thinking and problem-solving skills essential for the workforce. He emphasized that students must understand and use AI actively, not just consume it. He called on the committee to prioritize AI in K–12 to strengthen both career readiness and national competitiveness.

• Dr. Rafal-Baer noted that while AI is already impacting education, most schools lack resources for safe implementation. She recommended limiting federal mandates, strengthening data protections, and funding AI research. She also called for a White House Summit to guide national AI strategy. She emphasized that thoughtful, inclusive AI adoption can significantly improve education and workforce outcomes.

• Ms. Mote described AI as a transformative force that is unevenly adopted in schools. She stressed the need for equitable access, data safeguards, and educator support. Ms. Mote warned that relying on state-by-state policies would create fragmentation and called for federal leadership. She advocated for national investment to ensure AI is used safely and effectively in all communities.

• Mr. Chism emphasized workforce development and efficiency as key to AI’s role in K–12. He shared how AI tools can ease teacher workloads, citing a story where AI helped a veteran teacher stay in the profession. He argued that AI enhances, not replaces, educators and urged swift action to prepare students for an AI-driven economy.

SUMMARY OF KEY Q&A

• Rep. Rulli asked how AI helps students excel, calm parental concerns, and support students with disabilities. Mr. Chism explained that AI empowers high school students ethically, assists special education planning through custom plans, and encourages daily writing with teacher support, while advising parents to stay involved and informed.

• Rep. Bonamici asked how dismantling the Department of Education and cutting professional development programs would affect AI use in schools. Ms. Mote emphasized the Department's crucial role in tech guidance, data security, and professional development, while advocating for teacher AI literacy and federal infrastructure support. Dr. Rafal-Baer added that federal research provides valuable insight on effective AI tools and helps states and districts make informed decisions.

• Rep. Thompson asked how AI can help students with disabilities and what risks should be considered. Dr. Rafal-Baer said AI tools like screeners and tutoring aids can enhance special education and support less-experienced educators but warned that more research is needed to monitor outcomes and risks.

• Rep. Mannion asked about AI’s impact on student mental health and emotional development. Ms. Mote noted that while AI can aid mental health, strong human relationships have far greater impact and must remain central to student development.

• Rep. Messmer asked how states can support AI adoption and whether current guidelines are effective. Dr. Rafal-Baer said states should collect data, vet tools through “AI assurance labs,” and engage communities. Dr. Dobrin added that most state AI policies are still general, with limited collaboration, and emphasized the need for deeper research before widespread implementation.

• Rep. Hayes asked how eliminating the Office of EdTech and Department of Education impacts AI deployment. Ms. Mote responded that federal support is vital for cybersecurity, training, and guiding schools lacking internal AI expertise.

• Rep. Lee raised concerns about AI surveillance reinforcing systemic bias and asked whether companies check their tools for racial bias. Ms. Mote explained that bias is inherent in tools trained on historical data and emphasized the need for better federal data infrastructure and mitigation strategies. She said companies need broader datasets and collaborative oversight to address these issues effectively.

• Chairman Kiley asked the panel to expand on AI as a transformational “arrival technology.” Ms. Mote compared AI to electricity and urged collaboration, guardrails, and teacher training. Dr. Rafal-Baer warned that AGI is approaching and called for federal cybersecurity support. Mr. Chism emphasized researching teacher efficacy improvements. Dr. Dobrin suggested studying the impact of access disparities between free and paid AI tools.

• Rep. Wilson asked whether AI can function without human input and what happens when oversight is lost. Ms. Mote said AI needs human-trained data and that dismantling federal support harms families and teachers who rely on open data and expertise.
  Rep. Wilson asked how weakening civil rights protections affects AI bias and how AI could avoid deepening racial inequalities. Ms. Mote replied that educators need training and research to choose fair tools and address achievement gaps. She urged investments in digital infrastructure and targeted access programs to reduce bias.

• Rep. Harris asked how to prevent biased AI in schools and ensure content neutrality. Mr. Chism explained his district trains its own models on local servers and curates data sources to reduce bias.

  Rep. Harris asked about overregulation concerns. Dr. Rafal-Baer said federal mandates should focus on cybersecurity and privacy, not curriculum or literacy definitions.

• Rep. Adams asked how AI tools can support equity and prevent reinforcing bias. Ms. Mote emphasized AI literacy for parents, teachers, and students as key to navigating and questioning outputs.

  Rep. Adams also raised concerns about DEI funding cuts and diverse teacher recruitment. Ms. Mote stressed that teacher diversity and human connection are crucial to student success and AI should support—not replace—that.

  Rep. Adams asked about cuts to IES research. Ms. Mote warned that losing such data hinders schools' ability to make informed, equitable AI choices.

• Rep. Mackenzie asked how to ensure AI tools are truth-seeking and publicly vetted. Dr. Rafal-Baer recommended public state-led “AI assurance labs” for testing tools and promoting transparency. Dr. Dobrin emphasized demystifying AI for educators and called for tools built specifically for education, not retrofitted from industry.

• Rep. Miller asked how to protect student privacy and prevent AI misuse outside of school. Mr. Chism said his district scrubs data or uses a private server to avoid FERPA violations, and advised in-class projects to limit reliance on AI at home.
  Rep. Miller asked about AI’s role in special education. Dr. Dobrin said AI enables customized learning paths and real-time assessments, benefiting students with diverse needs.

• Rep. Scott asked if AI enhances teachers and whether it can support assessment and reduce bias. Ms. Mote said AI enhances, not replaces, teachers, but assessments must include human oversight to avoid flagging biases.
  Rep. Scott asked about costs and barriers to equitable adoption. Ms. Mote said costs include hardware and training, and stressed the need for industry collaboration and national policy alignment.

• Rep. Moylan asked what AI skills employers need and how to teach them. Dr. Rafal-Baer said employers want students to understand AI's risks, limitations, and ethics. Dr. Dobrin recommended aligning curriculum with industry use cases and focusing on AI competencies and fluency.

ADD TO THE NIMITZ NETWORK

Know someone else who would enjoy our updates? Feel free to forward them this email and have them subscribe here.