When we think about AI, we think about Silicon Valley or MIT labs because we are familiar with research innovated from these places. However, in a few years, this may change as China announced that they will become the world leader in AI by 2030.
Statistics show that the registered patent for AI in China during 2010-2014 increased three folds when compared to 5 years ago and the Chinese government hopes that this trend will continue to increase at an even faster pace.
The Chinese government views that AI is the key to economic development. It is estimated that the AI industry will amount to 1 trillion RMB (or around 5 trillion THB). In July, the Chinese government issued a three-tier roadmap that will take China towards this goal.
The first tier which begins now until 2020 is dedicated to “setting up the foundation”. This includes nurturing personnel in AI, basic software development to support AI, and setting the proper regulations, standards, and ethics for AI. At the stage, China hopes that their companies and research institutes will rise to the US standard.
The next tier is until 2025 wherein China aims for a “major breakthrough” in AI for the industrial and economic sector. The final stage is in 2025-2030 which is China’s final sprint to the finish line. China will become the “world’s AI research hub” and “leader in the new economic system”.
Today, we are aware of giant Chinese tech firms like Alibaba and Baidu who own competitive AI technologies. Yet the interesting question is what other elements will contribute to China’s AI superiority?
Bloomberg has three suggestions and the first two involves the “amount” of the population. With it, whenever China moves the world trembles. With China’s dedication to AI, the number of Chinese software engineers becomes an advantage. Moreover, China’s 751 million internet users is a significant amount of people to efficiently test just about anything.
The third reason may only be what companies in western countries dream of, it is government support. This is not support in terms of funding or tax policies but is support of “open data” for AI development. Such “openness” may be unimaginable in western countries due to different privacy concerns.
An example is with companies such as SenseTime who has developed AI that can recognize objects and faces. They had access to VDO footage from the police department in Guangzhou for AI development. This is significant as Guangzhou has
14 million people. SenseTime executive states that “China has a large population so collecting data is very easy. And when we talk about data, the government has the most resources”.
Such “open data” by the government becomes an advantage that companies in western countries can hardly access. These data include health, finance, and various other personal information. Looking back, there once was a collaboration between a public health organization in the UK who shared data with Google’s DeepMind lab for AI development. They were later sued by a privacy protection organization for violating personal privacy rights. Such difficulty rarely happens in China.
Nevertheless, some (especially western analysts) worry about the advancement of Chinese AI. They are concerned about the data protectionism policy as China’s cybersecurity law states that foreign companies need to store Chinese citizens’ information on domestic servers. They are not allowed to share the data collected in China to other service providers (The Economist questions whether this policy is implemented upon Chinese companies).
Additionally, concerns also involve ethics and preventive measures for AI such as boxing. Some analysts worry that China still doesn’t have such precautions. Moreover, concerns extend to the development of autonomous weapons as Chinese researchers haven’t signed the 2015 Open Letter which opposes the matter.
Another issue is the Chinese government’s standpoint regarding surveillance. China is looking to implement “social credit scores” to isolate those who “obey” from those who “oppose” the government. This is done by observing people’s behavior in the actual and social environment such as using inappropriate words on social media and expressing inappropriate behavior towards state officers.
The increased intelligence of AI, especially those originating from companies with good relationships with the government, may make control over the public tighter than ever.
It is worth witnessing whether China’s attempt will succeed by 2030 and what will be sacrificed along the way.