VC Hans Tung: China Outbuilds America, and Compute Is the New Railroad

By Grace Kim

One Chinese state conglomerate builds more commercial ships every year than the entire U.S. shipbuilding industry has produced in the 81 years since World War II ended. That figure sits at the center of Hans Tung's argument that the AI race will be decided by infrastructure, not ideas.

Tung, managing partner at Notable Capital and an early backer of Airbnb, Coinbase, and Anthropic, laid out the case in an interview with Fortune published October 13, 2026. His frame: take the venture-industry cliché that AI is an "Industrial Revolution-level change" and actually follow it through history.

He started with ships.

What did steel ships teach us about tech dominance?

"The Dutch, the Spanish, the Portuguese—they sailed around the world with beautiful large ships, but they were wooden wind ships," Tung said. "The British were the first to scale ships with the steam engine. Adopting steel technology and applying it to the steam engine—that made the British Navy stronger than the rest of the world combined."

In 1890, 80% of the world's ships came from British shipyards. Tung's read: a major technology only translates into geopolitical dominance when the infrastructure exists to back it up.

"Wood ships don't last as well in the long run," Tung said. "They end up dominating the world for more than 100 years."

His takeaway in plain words: stop fixating on any single app or company. The buildout is the story.

Are data centers the railroads of the 21st century?

Tung's answer is essentially yes. His proof point is the American Civil War, which the North won partly because its railroad network dwarfed the South's.

"Most major Civil War battles happened within 24 miles of a railroad… If you don't have a railroad—or compute—it doesn't matter if you have better engineers, better politicians, or better thinkers," Tung said.

"If you don't have enough compute to meet demand, it doesn't matter if you have a better model," he told Fortune. "The models that win are the ones that are available."

The point for AI: distribution capacity beats raw invention.

How far ahead is China, really?

Tung, who is Taiwanese-American, flagged China's construction record as the uncomfortable benchmark. Over the last two decades, China has built subway and high-speed rail at a pace he calls unprecedented.

"Now, China has more high-speed rail than the rest of the world combined, kind of like the North in the Civil War," Tung said. "They executed very, very well. It's mind-boggling."

The shipbuilding numbers sharpen the picture. By some estimates, over 80% of new shipbuilding orders worldwide run through China.

Compute is where the U.S. still leads—for now:

• China controls roughly 24 gigawatts of working data center compute capacity.
• The U.S. holds at minimum 47 gigawatts by conservative estimates.
• A Columbia Business School professor estimates the AI infrastructure buildout could average 3.6% of U.S. GDP every year through 2032.

That buildout carries its own risks, from community pushback to questions about whether the electric grid can handle the load.

What's the bottom line for investors?

Tung's conclusion cuts against standard VC orthodoxy. The best ideas and the best talent don't win on their own. In a technology-driven socioeconomic shift, invention matters less than having the infrastructure to deploy AI at scale.

"We're not going to just win," Tung said. "We have to work hard and make sure we win."

For a venture capitalist whose portfolio includes Anthropic, that's a striking stance: the model layer he backed may matter less than the gigawatts underneath it.