Energy, Minerals, and the Physical Stack Behind AI
Summary
A discussion with Turner Caldwell of Mariana Minerals and Drew Baglino of Heron Power highlights the critical physical infrastructure challenges underpinning America's AI economy and re-industrialization efforts. The US lags 50 years behind China in critical mineral supply and relies on grid infrastructure designed a century ago, while demand for both accelerates. Caldwell's Mariana Minerals is applying autonomous systems and reinforcement learning via Capital Project OS, Plant OS, and Mine OS to accelerate mining and refining projects, including a copper mine in Southeast Utah and a lithium refinery in Texas, aiming for 10 projects in 10 years. Baglino's Heron Power is developing solid-state transformers to replace aging mechanical grid equipment with silicon and software. Both emphasize lessons from Tesla, including techno-optimism and risk appetite, and advocate for durable industrial policy, smarter permitting, and a federal grid investment framework to boost US competitiveness and create hundreds of jobs.
Key takeaway
For Policy Makers and Entrepreneurs aiming to secure America's AI future and re-industrialize, recognize that dominance hinges on physical infrastructure, not just algorithms. Prioritize durable industrial policy, streamline permitting processes, and establish a federal grid investment framework. This approach will mobilize private capital, accelerate critical mineral capacity, and modernize the power grid, fostering domestic manufacturing and high-paying technical jobs.
Key insights
AI dominance and re-industrialization hinge on modernizing US critical mineral supply and grid infrastructure through technology and policy.
Principles
- Innovation can revitalize archaic industrial systems.
- Co-locating supply chains outweighs labor cost differentials.
- Mission-driven culture attracts top talent.
Method
Mariana Minerals uses Capital Project OS, Plant OS (reinforcement learning for refineries), and Mine OS (reinforcement learning for mining) to accelerate project delivery and increase autonomy in mineral operations.
In practice
- Apply reinforcement learning to optimize refinery operations.
- Replace mechanical grid equipment with solid-state transformers.
- Seek talent from analog industries for new manufacturing.
Topics
- Critical Minerals
- Grid Modernization
- Industrial Policy
- AI Infrastructure
- Supply Chain Reshoring
- Power Electronics
- Autonomous Mining
Best for: Computer Vision Engineer, Policy Maker, Entrepreneur, Investor
Related on AIssential
Editorial summary, takeaway, and curation by AIssential. Original article published by The a16z Show.