Earlier today, the Commerce Department’s Bureau of Industry and Security (BIS) updated last year’s chip controls targeting China. The changes are indicative of an emerging fusion of two American political preoccupations: U.S.-China competition and hype surrounding AI. If lawmakers let it, that convergence stands to unnecessarily escalate tensions between the two countries.
BIS’ updates are designed to close loopholes in the original October 7, 2022 export controls and ensure that companies like Nvidia, which created special chips to comply with the rules, cannot sell China advanced semiconductors.
The argument is that the chips could end up aiding China’s military modernization. According to a press release on the updated rules, China’s access to advanced semiconductors threatens U.S. national security because “they can be used to improve the speed and accuracy of military decision making, planning, and logistics.”
In a briefing ahead of the release of the updated rules, Commerce Secretary Gina Raimondo made her perception on the threats posed by both the People’s Republic of China and artificial intelligence very clear. She called AI “probably the most obvious example of the kind of transformational technology that we have to assess and control.” It “can do tremendous and profound harm if it’s in the wrong hands and in the wrong militaries,” she added.
China, like the United States, is developing military strategies that will utilize large language models (AI) and unmanned drones, among other AI applications, in an attempt to win the advantage in a race to end all races. In August, U.S. Deputy Defense Secretary Kathleen Hicks announced a new drone program to counter China’s military, following many similar initiatives aimed at building up military innovation. America’s advantage, according to Hicks, is its ability “to imagine, create and master the future character of warfare.” To that end, the Department of Defense requested $1.8 billion to use for AI in fiscal year 2024.
Many Americans will accept at face value that it is in the United States’ interest to prevent the People’s Liberation Army from obtaining AI-enabled weapons that could destroy the planet. But a concurrent consideration of Washington’s confrontational approach to the bilateral relationship and its nonexistent domestic AI regulations suggests that U.S. lawmakers have become intoxicated by the coalescing temptations of China hawkishness and AI hype.
AI hype in this context means giving credence to “doomers” who profess a sci-fi-esque vision of unhinged AI annihilating humanity—resulting in fear-induced paralysis. What’s the point of ensuring writers aren’t replaced by LLMs if we’re all likely to die at the hands of rogue robots anyway?
Emily Bender and Alex Hanna have argued that rather than give credence to these theories, which work to the immediate benefit of AI firms by distracting governments and the governed from companies’ current (highly profitable) capabilities, policymakers should instead pay attention to independent and peer reviewed scholarship on the risks of AI “and the harms caused by delegating authority to automated systems.”
The AI hype machine has clearly succeeded in obscuring several relatively immediate concerns—starting with how lawmakers can protect citizens from the non-hypothetical applications of AI that have been (not might start) destabilizing American schools, publications, and courts—and ending with how the U.S. plans to dissuade itself, in addition to China, from developing AI-enabled weapons of mass destruction.
For better or worse, foreign policy rarely begins with a consideration of human wellbeing over national interest. That is an unfortunate, if contextually admissible, reality. But it would follow the administration’s own logic for its next move to be appealing for robust domestic data and AI regulation, setting an example and creating standards for responsible civilian development of the technology, rather than expanding its already significant chip restrictions on China. As Treasury Secretary Janet Yellen has said, “credibility abroad begins with credibility at home.” Yellen also described U.S. economic strategy as “centered around investing in ourselves.”
There are many ways to invest in ourselves. If U.S. lawmakers had the informed confidence necessary to effectively regulate domestically-developed advanced AI—which relies on the chips the updated rules make sure China is cut off from—our export controls would at least have stronger legs to stand on.
The reality is conceptualizing and enacting AI regulation rooted in reality, not hype, is a more cutting edge area of global leadership than setting in motion yet another weapon that could end the world. We have been there and done that.
Read the full article here
