President Donald Trump signed an executive order on AI on Tuesday, aimed at reducing cybersecurity risks and establishing a voluntary government testing program for new models thirty days before public release. This order, a slimmed-down version, was initially delayed due to concerns about hindering U.S. competition with China.
However, the argument that U.S. AI regulation will cause it to fall behind China overlooks China's extensive and burdensome AI regulations over the past four years, during which it largely caught up with the United States in AI technology. China implemented over half a dozen binding national regulations since 2021, focusing on censorship, mandatory labeling of AI-generated content, mitigating psychological harms, and addressing AI-related job losses. These regulations, particularly the 2023 generative AI rule, require extensive vetting and testing, initially slowing deployment but ultimately leading to globally competitive models like DeepSeek V3 by early 2025. China's success, despite regulatory hurdles, is attributed to its strong technology ecosystem—access to capital, research talent, computing power—and its "small, fast, flexible" regulatory approach (_xiao, kuai, ling_). This iterative method, involving direct engagement with companies, has enhanced regulators' understanding and streamlined enforcement, proving regulation and competition are not incompatible.
No comments:
Post a Comment