AI Faces Energy Challenges
White House adviser David Sacks has his work cut out for him (POLITICO Pro’s ENERGYWIRE, subscription).
What’s going on: “Today, as the president’s AI and crypto czar, Sacks is tasked with opening doors for Silicon Valley and global financiers to expand AI infrastructure—and fast. But that presents Sacks with a wholly unfamiliar challenge: ensuring the tech industry gets the electricity it needs for a massive buildout of data centers to house the computing power needed for AI technology.”
- Technology giants including Amazon have pledged to spend billions on AI infrastructure to construct those data centers. In January, Trump and two tech firm leaders announced the $500 billion “Stargate” project to do just that.
Interconnection, cost challenges: While the president “has promised to use emergency declarations to build more power plants” to run these data centers, “tech companies that want to erect their own power plants will wrestle with the limited supply of ready transformers and the real-world challenges of putting up a large gas-fired power station or tying into a nuclear reactor.”
- That’s because regional electric grids and utility companies are already straining to keep up with Americans’ growing power appetite, and the prospect of dramatically increasing capacity is leading to concerns about cost and reliability.
What’s next: “Under Trump’s executive order, Sacks will work alongside policy staffers on science and technology, national security, economic policy and other relevant departments to craft an AI action plan. … Some direction could come from Trump’s National Energy Council, set to be led by Interior Secretary Doug Burgum. That council is likely to consider steps to streamline permitting and to unleash more fossil fuel production and other energy sources like nuclear and geothermal.”
What we’re doing: The NAM has long urged Congress to take steps to facilitate the construction of data centers, including:
- Reforming the U.S. permitting system;
- Ensuring energy affordability;
- Expediting licensing;
- Addressing shortages of components;
- Mitigating cybersecurity risks; and
- Bolstering workforce development efforts.