Practical AI AI policy and the battle for computing power
87 snips
Mar 9, 2026 
Ben Buchanan, Assistant Professor at Johns Hopkins SAIS and former White House AI advisor, brings a policy and geopolitical lens to computing power. He discusses why compute, not data, drives AI progress. He explores chip supply chains, Taiwan’s strategic role, government–industry relations, export controls, and the challenges of international AI governance.
AI Snips
Chapters
Books
Transcript
Episode notes
Chip Supply Chain Is A Geopolitical Leverage Point
- Semiconductor manufacturing is geopolitically concentrated, with TSMC in Taiwan producing the vast majority of advanced chips.
- Buchanan highlights ASML and an allied supply chain (US, Netherlands, Japan) as democratic strengths to protect.
Rebuild Domestic Chipmaking Capacity
- Use industrial policy to rebuild domestic chip capacity rather than rely solely on foreign suppliers.
- Buchanan cites the bipartisan Chips and Science Act and US fabs starting in Arizona as concrete steps taken.
AI Policy Has Had Bipartisan Foundations
- AI policy has generally been bipartisan and not inherently partisan, with both Trump and Biden-era actions on chip export controls and Dutch cooperation.
- Buchanan notes shifts in rhetoric but emphasizes underlying consensus on national security concerns.
