American credulity, American dynamism | Andreessen Horowitz
TLDR
- a16z partner argues AI is America’s next Manhattan Project, citing recurring pattern of panic-then-dominance across Sputnik, Japan’s rise, and the internet.
Key Takeaways
- Japan grew from 29th to 2nd largest economy 1950-1990; Harvard, pundits, and Trump predicted American defeat – the U.S. still ended up with 19 of the world’s top 25 companies.
- The U.S. spent 1.1% of GDP ($300B+ in today’s dollars) on internet infrastructure 1997-2001, seeding Netscape, AOL, Amazon, and Google.
- NASA doubled its budget six consecutive years after Sputnik, peaking at 4.4% of federal budget in 1966 and employing 400,000 people.
- Manhattan Project mobilized 130,000 workers across three parallel uranium enrichment pathways; Germany’s parallel effort peaked at roughly 30 people.
- Keil cites current AI-era examples: autonomous naval vessel welders in Louisiana, autonomous combat aircraft production in Ohio, AI-powered defense factories in Alabama and Arizona.
Why It Matters
- The piece is an explicit investment thesis framing AI as a geopolitical race America wins only if capital and belief mobilize at wartime scale.
- Historical analogies are load-bearing: each prior panic (Japan, internet) ended in U.S. structural dominance, and Keil uses that pattern to dismiss current AI skepticism.
- The argument lands at a specific policy claim – the bottleneck is belief, not capability – which shapes how a16z’s American Dynamism portfolio is positioned publicly.
Christian Keil, Partner at Andreessen Horowitz · 2026-04-08 · Read the original