Trump vows 3-week nuke permits, calls Greenland 'big ice'

At Davos, President Trump promised to fast-track nuclear power plant permits for datacenters to just three weeks, while also discussing AI dominance and making geopolitical comments about Greenland. Nvidia's Jensen Huang separately presented his vision of AI as a five-layer cake, emphasizing energy needs and global adoption.

At the World Economic Forum in Davos today, President Donald Trump delivered a wide-ranging speech that touched on nuclear energy permitting, AI leadership, and geopolitical tensions, while Nvidia CEO Jensen Huang presented a technical vision of AI's future structure. The events highlighted the growing intersection of energy policy, national security, and artificial intelligence development.

Fast-Track Nuclear Permits for Datacenters

Trump announced a dramatic acceleration of the permitting process for nuclear power plants, specifically targeting datacenter development. "We're leading the world in AI by a lot. We're leading China by a lot," Trump claimed, attributing this leadership to his administration's approach to energy infrastructure.

The president described a meeting with technology executives where he proposed a solution to the grid constraints facing datacenter expansion. "I came up with the idea. 'You know, you people are brilliant. You have a lot of money. Let's see what you can do. You can build your own electric generating plants.'" Trump told the audience that he promised tech companies, "Not only am I not kidding, you're going to have your approvals within two weeks" for electricity generation, with "Nuclear will take three weeks."

This represents a significant departure from current regulatory timelines. In the United States, the Nuclear Regulatory Commission typically requires 4-6 years for new reactor licensing, with construction and operational permits extending even longer. Trump's proposed three-week timeline would represent an unprecedented acceleration, though the legal mechanisms for such rapid approval remain unclear.

The president contrasted this approach with European energy policy, criticizing what he called the "radical left" influence. He noted that "The United Kingdom produces just one third of the total energy from all sources that it did in 1999," despite having access to "one of the greatest reserves anywhere in the world" in the North Sea.

Geopolitical Comments on Greenland

Trump addressed speculation about potential U.S. acquisition of Greenland, which he described as a "big, beautiful piece of ice." He stated, "We won't get anything unless I use excessive strength and force, when we would be unstoppable. I won't use force." However, he added a warning: "[It's] very important we use it for national and international security," and "We'll see what happens."

These comments follow previous statements from the Trump administration expressing interest in Greenland's strategic value, particularly for Arctic positioning and resource access. Greenland remains an autonomous territory within the Kingdom of Denmark, and both Denmark and Greenland have previously rejected any notion of sale or transfer.

Nvidia's Five-Layer AI Framework

Nvidia CEO Jensen Huang presented a technical framework for understanding AI development, describing it as a "five-layer cake" with energy needs forming the foundational layer and value creation occurring at the application layer. This model emphasizes that AI infrastructure requires substantial energy inputs before producing meaningful outputs.

Huang pushed back against suggestions of an AI bubble, pointing to practical constraints in the market. "If you try to rent an Nvidia GPU in these days, it's so incredibly hard, and the spot price of GPU rentals is going up, not the latest generation, but two generation old GPUs," he explained, suggesting demand continues to outpace supply.

The Nvidia CEO emphasized that AI is increasing rather than destroying demand for skilled labor, particularly in chip manufacturing, power infrastructure, and datacenter construction. He noted that Europe may be better positioned in terms of "trades" compared to the United States, while AI helps address skills shortages in fields like medicine by reducing administrative burdens for nurses and accelerating radiologist diagnoses.

Global AI Adoption and Open Models

Huang advocated for international participation in AI development, stating, "I really believe that every country should get involved to build AI infrastructure, build your own AI, take advantage of your fundamental natural resources, which is your language and culture." This contrasts with more protectionist approaches that focus on domestic AI dominance.

The open model approach allows countries to develop AI systems tailored to their specific languages and cultural contexts, potentially reducing dependency on foreign technology while preserving linguistic diversity in AI applications.

Export Controls and National Security Concerns

The Davos discussions revealed tensions within the AI industry regarding export controls. Anthropic CEO Dario Amodei recently criticized the U.S. government's decision to approve exports of H200 chips and AMD parts to China, calling it "a big mistake" with "incredible national security implications."

Amodei compared the situation to "selling nuclear weapons to North Korea and [bragging that] Boeing made the casings," highlighting concerns that advanced AI chips could enhance China's military capabilities despite U.S. technological advantages in silicon development.

Regulatory and Compliance Implications

The proposed three-week nuclear permitting timeline raises significant regulatory questions. The Nuclear Regulatory Commission operates under strict safety protocols established by the Atomic Energy Act of 1954 and subsequent amendments. Any acceleration would require substantial changes to existing frameworks, potentially facing legal challenges from environmental groups and safety advocates.

Datacenter operators currently face compliance requirements under multiple regulatory regimes, including environmental impact assessments, grid interconnection studies, and safety reviews. Nuclear facilities add layers of complexity involving radiation protection, waste management, and emergency preparedness planning.

The International Atomic Energy Agency provides guidelines for nuclear safety that influence domestic regulations. While the U.S. maintains its own standards, rapid permitting could create tensions with international best practices and potentially affect the country's standing in global nuclear cooperation agreements.

Energy Infrastructure Challenges

The discussion underscores the growing energy demands of AI development. Training large language models requires substantial computational power, with estimates suggesting that datacenter energy consumption could double by 2026. This has led to increased interest in nuclear power as a baseload energy source that can provide consistent, carbon-free electricity.

However, nuclear construction faces significant challenges, including high upfront costs, long development timelines, and public opposition. Small modular reactors (SMRs) have been proposed as a more flexible alternative, but these technologies remain in early deployment stages.

Market Implications

The promise of rapid nuclear permitting could influence investment decisions in the datacenter sector. Companies considering expansion may factor in potential energy cost savings from on-site nuclear generation, though the practical implementation timeline remains uncertain.

GPU rental prices mentioned by Huang indicate continued supply constraints in the AI hardware market. The difficulty in accessing even older generation GPUs suggests sustained demand that may persist despite potential regulatory changes in energy infrastructure.

Looking Forward

The Davos discussions highlight the complex interplay between energy policy, technology development, and national security. While Trump's proposed nuclear permitting acceleration remains to be implemented, it signals a potential shift toward prioritizing energy infrastructure for AI development.

The contrasting approaches between Trump's focus on U.S. dominance and Huang's advocacy for global participation reflect broader debates about AI governance and international cooperation. As AI capabilities advance, the energy requirements and geopolitical implications will likely continue to shape policy discussions at forums like Davos.

For datacenter operators and AI companies, the evolving regulatory landscape presents both opportunities and uncertainties. The promise of faster energy infrastructure development could enable expansion, but the practical implementation of such policies remains to be seen.

#Nuclear Energy #AI_Infrastructure #Data Centers #Geopolitics #Export Controls