The United States is at risk of blowing its lead in the artificial intelligence (AI) technology race to China, a development that could have profound implications for national security and technological dominance, experts told the Daily Caller News Foundation.
China is making strides in incorporating AI technology into its military and is at the forefront of worldwide AI research, but the U.S. remains the leader in generative AI, such as popular chatbot ChatGPT, according to experts, studies and government documents. The main U.S. strategy to slow down China’s advancement through semiconductor (chip) restrictions is insufficient and while the race is not a foregone conclusion, the U.S. must urgently accelerate and adapt its approach if it wants to secure victory, experts told the DCNF.
“It’s no secret that the Chinese government wants to militarize its own AI systems and we are shaping up to be on the losing end of that digital arms race if we don’t get serious,” Joel Thayer, president of the Digital Progress Institute, told the DCNF.
China’s dedication to AI advancement is evident from its “New Generation Artificial Intelligence Development Plan” in 2017, which underscores AI integration into its military. “Promote all kinds of AI technology to become quickly embedded in the field of national defense innovation,” the plan states. “Strengthen the construction of military and civilian AI technology standard systems.”
In February, a Chinese AI pilot beat a human in a dogfight, which ended in about 90 seconds, South China Morning Post reported. China’s People’s Liberation Army (PLA) detailed a strategy in December 2021 about the incorporation of AI, asserting China planned to deploy a military in which “man and machine” engaged in combat side by side, according to a DCNF translation.
The PLA claimed it was working on an AI “superorganism” with “deep learning,” “evolution” and deception capabilities, according to the translation.
A man takes a picture of robots during the World Artificial Intelligence Conference (WAIC) in Shanghai on July 7, 2023. (Photo by WANG ZHAO/AFP via Getty Images)
Further, out of the 10 leading companies producing AI research, four are Chinese, according to a study by Nikkei Asia in collaboration with Dutch scientific publisher Elsevier. They are Tencent Holdings, Alibaba Group Holding and Huawei Technologies.
Anja Manuel, executive director of the Aspen Security Forum, told the DCNF that the U.S. undoubtedly continues to lead in generative AI and mentioned the argument that China’s strict regulations and authoritarian government may hinder the country’s innovation.
“Jury’s out on that one, but we do know that the Chinese government is very focused on making their companies successful in AI,” Manuel told the DCNF. “[It] has given … companies clear lanes about which tech innovation is just fine to pursue: for example, anything in visual recognition AI, autonomy (drones, cars) etc. Many Chinese companies are excellent at these specific AI applications.”
Beijing-based tech giant Baidu referenced China Science Daily’s reporting that its generative chatbot Ernie overtook ChatGPT — an older iteration of the chatbot — in benchmark tests and beat the advanced GPT-4 in Chinese language tests in June. However, Ernie’s “comprehensive ability is slightly inferior to GPT-4,” according to China Science Daily.
The U.S. has worked on decelerating China’s AI advancement by limiting its access to semiconductors and other equipment.
“Most technologists agree that export controls on chips etc — although an important part of the policy tool box — won’t alone keep the U.S. in the lead over China,” Manuel told the DCNF. “The bottom line is that the U.S. is innovative and still in the lead, but to stay there, our focus should be on running faster ourselves. We won’t get there just by trying to ‘hobble’ our competitor.”
Thayer agreed. “Simply relying on moderating the chip market is not enough,” he said.
Moreover, Microsoft has a China-based AI chatbot dubbed Xiaoice that has 660 million worldwide users, according to the company.
Jake Denton, a research associate at the Heritage Foundation’s Tech Policy Center, told the DCNF that while Microsoft is an egregious example of an American company’s AI connection to China, other tech companies’ presence in the region is concerning as well. “There’s a lot of interconnectivity between the offices in Asia and North America,” he said. “Regardless of if it’s a formal office established purely for AI research or a generalist office, there are still implications. Any Silicon Valley company with a presence in China is going to in some way, be trying to please China.”
Denton raised the example of Apple having a multibillion-dollar deal with China and the company disabling the airdrop feature during Hong Kong protests against the Chinese government. “You have to assume the same exact requests would be granted for AI development,” Denton said. “That’s just the logical progression of this. All those companies, they’ll just shut you out of China if you don’t comply.”
Thayer also said the U.S. cannot rely on tech companies in this tech race.
“The AI market is concentrated among a few companies, dominated by Big Tech,” Thayer told the DCNF. “Worse, almost all have significant ties to China and its government. We can’t depend on industry alone to remove this threat China poses given certain market realities, especially in the tech sector.”
“We can still win this, but we need to turn this ship around!” Thayer said.
Scientist and AI entrepreneur Gary Marcus told the DCNF that the outcome of the tech race is undecided, but it is creating too much tension and distance.
“A lot of the emphasis has been on the race between China and the US – which is certainly not decided yet – but we need to dial down the temperature a bit, as collaboration between the countries is going to be vital, unless and until we can figure out how to control AI and mitigate the short and long-term risks that both countries face.”
Post a Comment