miércoles, 3 de enero de 2018

China: ¿Smart killer robots?


A recent report by Accenture Technologies suggests that by 2035 Artificial Intelligence (AI) might add as much as 1.6 percentage points to China’s Gross Domestic Product (GDP).


The researchers at Accenture argue that to yield maximum economic benefits from AI, China needs to use it as an additional factor of production, relying on AI for intelligent automation, labor and capital augmentation and innovation diffusion.


Analysts from other institutions also support such claim: PricewaterhouseCoopers projected that by 2030 global GDP could increase by $15.7 trillion, with almost half of these gains coming from China. PwC estimates that AI will account for a 26 percent GDP boost, or $7 trillion, in China in the next 13 years.


However optimistic these economic projections sound, rapid development of AI in the PRC might also bring about significant challenges to freedom and security, both in China and beyond: Chinese companies are barely constrained by legal issues over data collection and users’ privacy. They can freely use the data of China’s 750 million internet users, match photos and personal IDs and then train facial recognition algorithms and other types of neural networks on large datasets.


Currently the Chinese government not only does not restrict the uses of personal data by domestic tech giants such as Tencent, Baidu and Alibaba, but actively uses their resources to surveil Chinese citizens: Beijing has successfully used advanced facial recognition systems to analyze data from millions of cameras to track down law violators. On the one hand, this might help police to catch criminals faster. But on the other hand, and bearing in mind that China is more rapidly than any other country moving towards becoming a surveillance state where no one can hide from the government, the AI can also be used to censor online environments and identify people who spread “sensitive” information online. Through the use of the AI, any news considered "not good enough" for the government could be censored automatically as it gains steam, and then relaxed after the storm has passed. And there would be almost no trace that it had ever happened. Applying these techniques to the internal affairs, the possibility of implanting an Smart Dictatorship could be enormous.


Regarding military uses of AI, the U.S. is already alarmed about China potentially using AI for that purposes: In February, Elsa B. Kania, an analyst with The Long Term Strategy Group at the Aspen Institute, testified before the U.S.-China Economic and Security Review Commission that the Chinese People’s Liberation Army has reportedly begun incorporating AI into unmanned weapon systems, including UAVs and UUVs: “The Chinese defense industry has likewise made significant progress in its research and development of a range of cutting-edge unmanned systems, including those with supersonic, stealth, and swarming capabilities, but appears to face continued challenges in UAV engines, data links, and sensors,” she said. Of course, U.S. still has an advantage in terms of AI development, but China is catching up fast: Given the amount of resources available in China for the AI development, chances are Beijing will be able to fulfill its goal of becoming the industry leader by 2030.