China lags behind the US when it comes to the artificial intelligence that powers chatbots like ChatGPT. However, China is ahead of the curve when it comes to producing scientists who will support a new generation of humanoid technology.
China has overtaken the United States on several metrics to become the largest producer of AI talent, with China producing nearly half of the world's top AI researchers, according to a new study. By contrast, about 18 percent come from U.S. undergraduate institutions, according to the study, and MacroPolo, a think tank run by the Paulson Institute that promotes constructive relations between the United States and China.
The findings show a sharp increase in China, which three years ago produced about a third of the world's top talent. In contrast, the United States remained about the same. The study builds on the biographies of researchers whose papers were presented at the 2022 Neural Information Processing Systems Conference. NeurIPS, as it is known, focuses on the advances in neural networks that have underpinned recent developments in generative AI.
Talent imbalances have been building up for much of the past decade. For much of the 2010s, the United States benefited as many of China's top talents moved to American universities to earn their Ph.D. The majority of them stayed in the United States. However, this survey shows that trends are starting to change, with an increasing number of Chinese researchers in China.
What happens in the next few years could be crucial as China and the US vie for dominance in AI, a technology that has the potential to improve productivity, strengthen industries and drive innovation. researchers could become one of the most geopolitically important groups in the world.
Generative AI has captivated Silicon Valley and China's tech industry, sparking a frenzy in funding and investment. This boom is being driven by US tech giants like Google and startups like OpenAI. This could attract Chinese researchers, but rising tensions between China and the United States could deter some, experts said.
(The New York Times sued OpenAI and Microsoft for copyright infringement of news content related to AI systems.)
One of the reasons why China has been able to cultivate so many AI talents is because it has invested heavily in AI education. Makropolo managing director Damian Marr said that since 2018, the country has added more than 2,000 undergraduate AI programs, with more than 300 at the most elite universities, but that the programs are a breakthrough. He pointed out that there was not much emphasis on the technology that brought about this. By chatbots like ChatGPT.
“Many of the programs are about AI applications in industry and manufacturing, not about the generative AI that has now come to dominate the American AI industry,” he said.
The United States has been at the forefront of breakthrough advances in AI, most recently with the incredible human-like abilities of chatbots, much of the work done by Chinese-educated researchers.
According to the survey, researchers from China currently account for 38% of the top AI researchers working in the United States, of which 37% are Americans. Three years ago, 27% of the top talent working in the United States came from China, compared to 31% from the United States.
“The data shows how important Chinese-born researchers are to America's AI competitiveness,” said Matt Sheehan, a fellow at the Carnegie Endowment for International Peace who studies Chinese AI.
He added that the data seems to show that the US remains attractive. “We are a world leader in AI because we continue to attract and retain talent from all over the world, especially from China,” he said.
Peter Abbeer, a professor at the University of California, Berkeley and founder of AI and robotics startup Covariant, works alongside a number of Chinese researchers inside major U.S. companies and universities. He said it was natural.
“It’s just a natural situation,” he said.
Until now, U.S. defense officials had not been too concerned about an exodus of AI talent from China, in part because many of the largest AI projects did not handle sensitive data and Partly because it was better to have the best brain possible. Concerns were also fueled by the fact that much of the major research on AI is publicly available.
The Trump administration has introduced a ban banning students from some Chinese military-affiliated universities from entering the United States, and the influx of Chinese students into China has slowed down relatively due to the impact of the new coronavirus infection. Nevertheless, research has shown that a number of the most promising AIs exist. They continued to commute to the United States to study.
But a federal indictment alleges this month that a Chinese former Google engineer was charged with trying to transfer AI technology, including a critical microchip architecture, to a Beijing-based company for which he was secretly paid. was indicted for.
The significant number of Chinese AI researchers currently working in the U.S. will not impede the continued flow of top Chinese computer engineers to the U.S., according to experts focused on U.S. competitiveness. This poses a challenge for policymakers seeking to counter Chinese espionage.
“Chinese academics are pretty much at the forefront of the AI field,” said Subbarao Kambampati, a professor and AI researcher at Arizona State University. If policymakers try to ban Chinese people from doing research in the United States, they “shoot themselves in the foot,” he said.
U.S. policymakers have had a mixed track record. Trump administration policies aimed at curbing Chinese industrial espionage and intellectual property theft have since been criticized for falsely indicting numerous professors. Chinese immigrants say the programs encourage some to remain in China.
For now, the study found that most Chinese people who earned their Ph.D.s in the U.S. remain in the U.S., helping the U.S. become the global center of the AI world. Still, the US' lead is starting to decline, hosting about 42% of the world's top talent, down from about 59% three years ago, according to the survey.