Opinion | 3 low-hanging fruit for US-China cooperation on AI
Given the escalation of technological competition across the Pacific, it is tempting to see AI as an area where excessive caution and hype over security and politico-ideological control render it impossible for the two administrations to see eye to eye.
Yet this view ignores the vast areas in which Beijing and Washington can still work together through government-to-government and people-to-people engagement.
Even while comprehensive agreements on the regulation of dual-use AI remain elusive – given domestic political intransigence, high strategic stakes and the US’ ratcheting up of pressure to contain China’s technological rise – both administrations can pragmatically cultivate a modus vivendi.
China and the US exchanged views “deeply, professionally and constructively” in their first intergovernmental talks on AI risks and governance in Geneva on May 14, according to a Chinese foreign ministry readout. Photo: Weibo
In San Francisco last November, Chinese President Xi Jinping and US President Joe Biden affirmed the importance of coordinating over the challenges AI poses to humanity. The next step should be to address three low-hanging fruit when it comes to cooperation.
The first area concerns the potential detrimental effect of AI-induced automation on societies. The displacement of routine service labour and acceleration of substitution through AI-powered robotics could result in mass unemployment and widen socioeconomic inequality.
Experts argue that addressing AI-induced malaise – ranging from misinformation via deepfakes and algorithmic targeting to disruption of relationships and attitudes of trust and compassion – requires joint efforts from some of the world’s largest tech regulators, in China, the United States and the European Union.
While China’s extensive regulations on AI use in content generation and social media could prove to be unduly stifling for private developers, and granted that the US has a vastly different political economy, Washington may still benefit from taking a leaf out of China’s book on delineating harmful and impermissible corporate uses of user data. Aligning on transnational privacy and data regulation is of critical importance.
Secondly, it would take explicit instructions from the top to establish guard rails on the deployment of lethal AI-powered technologies in military settings. As the late US secretary of state Henry Kissinger noted, the gargantuan costs of an unbridled AI-powered war are far too daunting for responsible statesmen to ignore.
02:11
Autonomous ‘wingman’ drone technology being developed for tomorrow’s ‘Top Gun’ fighter pilots
Autonomous ‘wingman’ drone technology being developed for tomorrow’s ‘Top Gun’ fighter pilots
More substantively, both sides must commit to withholding decision-making powers from AI in the use of large-scale weaponry and in digital warfare against financial and physical infrastructure, as well as agree on heightened transparency and full disclosure of AI militarisation capacities.
Many of these steps had been adopted previously in nuclear non-proliferation and coordination between the Soviet Union and the US during the Cold War. We live in times when the focal bilateral relationship is much less fractious.
Indeed, China’s defence ministry has indicated its openness to jointly regulating military AI, and the recent US-China talks on the sidelines of the Shangri-La Dialogue in Singapore are a reassuring sign of thawing military-to-military relations.
Recently, the reported use of algorithms to identify human targets in the Gaza war has sparked considerable controversy. The use of AI in the Ukraine war has also raised concern.
In defining the ethical boundaries of war, legal and ethical experts should put their heads together to foster mutually binding regulatory standards, taking into full account the complexity of human experience and contextually specific expectations.
US Secretary of Defence Lloyd Austin shakes hands with China’s Defence Minister Dong Jun during a ministerial round-table discussion on the sidelines of the IISS Shangri-La Dialogue in Singapore on June 1. Photo: Bloomberg
Thirdly, the issue of non-alignment between humanity and AI knows no nationality or politics. As leading Chinese AI ethicist Yi Zeng noted in 2019, value alignment with humanity is a universal challenge for AI.
One classic example is philosopher Nick Bostrom’s 2003 thought experiment – suppose an AI agent has been tasked with producing as many paper clips as possible. In seeking to maximise the space, energy and resources for manufacturing, the powerful agent acts to stifle the growth of humanity, which it views as contrary to its production objective.
This may seem fanciful, yet is by no means unrealistic. With sufficiently powerful AI agents, seemingly clear objectives could produce unintended, perverse consequences – an issue both Chinese and American developers must confront as they race to produce the most competitive AI models.
03:08
What if robots took over the world? One ‘imagines’ nightmare scenario
What if robots took over the world? One ‘imagines’ nightmare scenario
Additionally, specific training environments and positive reinforcement will spur certain emergent goals in AI, which – when transplanted to a non-test environment – could produce undesirable outcomes.
Both countries’ leaders share an overarching interest in ensuring their AI agents do not impinge on humanity’s interests, whether through intentional manipulation by unaccountable private actors, or faulty developmental processes by authorised researchers.
Thought leadership and research on this issue need not be zero-sum – both China and the US stand to gain from alignment in ethical tech governance.
Here Hong Kong can serve as a conduit for Track 1.5 and 2 dialogues on AI, which could feature a larger proportion of private-sector players than such dialogues in Beijing – especially from the US and the rest of the West.
All hands must be on deck to avoid the calamities that could result from increasingly intertwined developments in nascent technologies and geopolitics.
Brian Wong is an assistant professor in philosophy at the University of Hong Kong, and a Rhodes Scholar and adviser on strategy for the Oxford Global Society
IE 11 is not supported. For an optimal experience visit our site on another browser. Trump Wraps China Summit With Xi Jinping: What Are the Results? 05:41 Now Playing Xi gives Trump rare tour of secret garden at heart of Chinese government 01:04 UP NEXT Trump departs China after two-day summit 01:01 Special Report: Trump
Expert breaks down China’s influence on Iran, US foreign policy Dennis Citrinowicz, an analyst at the Institute for National Security Studies, explains China’s complex role in the Iran conflict and its strategic implications for America. He highlights that China, while not wanting Iranian nuclear weapons, benefits from US focus on Iran, potentially draining munitions needed
Before leaving a two-day summit in Beijing, President Donald Trump said he had made many trade deals with China, accompanied by a cohort of “brilliant” tech billionaires. But the details of those deals – at least on Friday afternoon in Beijing when Air Force One departed – were vague, signaling a potential shift in leverage
Spiked fuel prices continue to rattle the world economy as talks between the United States and Iran to find a resolution to the war appear to have no end in sight. Data released by the Labor Department on Tuesday showed gas prices jumped 5.4% with average regular gas costing more than $4.50 a gallon, according
IE 11 is not supported. For an optimal experience visit our site on another browser. Trump Wraps China Summit With Xi Jinping: What Are the Results? 05:41 Xi gives Trump rare tour of secret garden at heart of Chinese government 01:04 Now Playing Trump departs China after two-day summit 01:01 UP NEXT Special Report: Trump
Cerebras, an AI chipmaker, saw its shares nearly double on Nasdaq, closing up 70% with a $95B market cap. Cerebras’s powerful chips are key in the US-China AI tech race. Chris Buskirk, co-founder and chief investment officer of 1789 Capital, a key Cerebras investor, says the company’s IPO is geopolitically significant. On Thursday, shares of
