“Data is what drives A.I., and China has more data than anyone else,” Kai-Fu Lee, CEO of Chinese venture capital firm Sinovation ­Ventures and author of AI Superpowers: China, Silicon Valley, and the New World Order, said at the Fortune Global Forum on Tuesday.

The world’s rapid shift to online platforms during the pandemic is a windfall for artificial intelligence, he said. The surge in online activity is building a mountain of data that companies and researchers can use to improve A.I. algorithms. “Data is everything to A.I. More digitization and more online [users] mean more data. And that’s more A.I. that could be trained,” Lee said.

During the COVID-19 pandemic, the U.S. has made digital gains in areas like food delivery and mobile payments, Lee said, but China has surged ahead in deploying new A.I. technologies like universal contact-tracing apps and robots that conduct automated temperature checks, deliver packages, and spray disinfectant. “China has a tenacious entrepreneurial ecosystem that has built companies that can harness the data,” he said.

A few years ago, Lee believed that China and the U.S. could cooperate—instead of compete—in artificial intelligence. But now, amid worsening U.S.-China relations and talk of “decoupling” the two economies, collaboration seems to be off the table.

“I had hoped there would be two superpowers with different strengths—the U.S. more on research, and China more on implementation—with ways to work together,” Lee said. “[Today] it looks like that dream is more difficult.”

Another trouble spot in the digitization trend is the mounting human cost.

As the pandemic accelerates the development of A.I., society has less time to figure out how to offset the cost to workers, Lee said. Software is increasingly able to replicate or exceed the performance of human laborers, just as companies are looking to cut costs amid a global economic downturn, which puts existing jobs at risk, Lee said.

Recent developments in A.I. and robotics make jobs involving customer service or factory labor especially vulnerable to automation, Lee said. “On the offline blue-collar work side, we find it’s just easier to put [artificial intelligence] in the middle of workers.”

As a result, employers and governments need to consider how A.I. may upend the lives of workers around the world, he said.

“We should really think [about] society and the world—what happens with growing wealth inequality and job displacements,” Lee said.