A new report from the Washington-based Center for a New American Security raised the alert level of the US defence community over China’s rise as an artificial intelligence (AI) superpower, one that could effectively destroy the American military by 2030.
The meticulous report no doubt will send a chill through the halls of the Pentagon.
The report, ‘Battlefield Singularity: Artificial Revolution, and China’s Future Military Power’, by Elsa Kania, paints a disturbing picture of China’s AI military modernisation programmes. Kania, as co-founder of the China Cyber and Intelligence Studies Institute, is well suited to write the investigative report using available Chinese-language open-source materials that reveal China’s military thinking and progress on AI.
Kania reported that China’s military is pursuing advances in ‘impact and disruptive military applications of AI’ and given it ‘high-level priority within China’s national agenda for military-civil fusion’. The goal is to become the world’s ‘premier innovation centre’ in AI by 2030.
According to the report, the Chinese military believes the advent of AI could fundamentally change the very character of warfare itself. Transforming itself from today’s ‘informatised’ warfare to ‘intelligentised’ warfare, in which AI will be critical to military power.
The result of this change would be the start of a major shift from China’s strategic approach, ‘beyond its traditional asymmetric focus on targeting US vulnerabilities to the offset-oriented pursuit of competition to innovate’.
China’s military is seeking ‘leapfrog development’ to achieve a ‘decisive edge’ in terms of ‘trump card weapons’ that prove a critical edge in ‘strategic frontline technologies’ against the US during a war. The report pointed out that the magnitude of Chinese publications in ‘deep learning’ has already exceeded the US as of 2014, and China ranks second in AI patent applications with 15,754 in total filed as of late 2016.
In July, China released the New-Generation AI Development Plan that articulated its ambition to lead the world in AI by 2030, becoming the premier global AI innovation centre. ‘Under this strategic framework, China will advance a three-dimensional agenda in AI: tackling key problems in research and development, pursuing a range of products and applications, and cultivating and expanding AI industry.’
This will include support for AI technologies that could result in paradigm shifts, including brain-inspired neural network architectures and quantum-accelerated machine learning. ‘The plan calls for building up national resources for indigenous innovation and pursuing continued advances in big data, swarm intelligence and human-machine hybrid intelligence…’
The report noted that Chinese teams dominated the ImageNet Large-Scale Visual Recognition Challenge, an AI computer vision contest, in 2016 and 2017. For the first time, at the 2017 annual meeting of the Association of the Advancement of Artificial Intelligence, China submitted an equal number of accepted papers compared to the US.
Then in November, Yitu Tech, a Chinese facial recognition start-up, took first place in the Facial Recognition Prize Challenge hosted by the Intelligence Advanced Projects Agency (IARPA). What the reader might find disturbing is that the Maryland-based IARPA is under the US Office of the Director of National Intelligence, which funds research across a range of technical areas, including mathematics, computer science, physics, chemistry, biology, neuroscience, linguistics, political science and cognitive psychology.
IARPA’s activities would be a natural fit for the Chinese Communist Party as it increases social control and stability through ‘new techniques for policing, censorship and surveillance, such as the installation of millions of surveillance cameras enhanced with AI technology’.
The report highlighted concerns the US should have on cooperative efforts with China. Chinese investments in Silicon Valley AI have fuelled the debate on whether the US Committee for Foreign Investment should expand reviews of Chinese high-tech investments, especially in AI. For example, the report pointed out the USAF became concerned after Chinese investment in Neurala, an AI start-up known for ‘innovative deep learning technology that can make more reactive robots’. The company is building the ‘Neurala Brain’ with a deep-learning neural network software.
Between 2012 and mid-2017, Chinese technology investments amounted to $19 billion in the US with particular focus on AI, robotics and augmented or virtual reality, said the report. In May 2014, Baidu Inc. established its Silicon Valley Artificial Intelligence Laboratory. In June 2014, Qihoo 360 Technology Co, a Chinese cybersecurity company, and Microsoft established a partnership in AI that focused on AI and mobile Internet.
In November 2015, the Chinese Academy of Sciences Institute of Automation (CASIA) and Dell established the Artificial Intelligence and Advanced Computing Joint Laboratory, which is pursuing development of cognitive systems and deep-learning technologies.
In January 2016, BEACON (Bio/computational Evolution in Action CONsortium), a centre located at Michigan State University, received funding from the US government via the National Science Foundation to establish the Joint Research Center of Evolutionary Intelligence and Robotics, headquartered at Shantou Technical University, also in partnership with the Guangdong Provincial Key Laboratory of Digital Signal and Image Processing.
In October 2016, Huawei Technologies devoted $1 million in funding to a new AI research partnership with the University of California, Berkeley. In April 2017 Tencent announced plans to open its first AI research centre in Seattle. That same month, Baidu Inc acquired xPerception, a US start-up specialising in computer vision.
The US is not the only accomplice. In 2011 and 2012, the University of Technology Sydney (UTS) established five research centres with Chinese universities that included centres on intelligent systems, data mining, quantum computation and AI. In 2017, UTS partnered with the China Electronics Technology Group (CETC) focusing on big data, AI and quantum technologies.
In 2014 Chinese drive-system maker Best Motion created a research and development centre at the University of Nottingham to develop high-quality servo drive systems for use in AI and robotics. In 2016 the Torch Innovation Precinct at the University of New South Wales was established as a joint China-Australia science and technology partnership to research military-relevant technologies, such as unmanned systems.
In March of this year the Hangzhou Wahaha Group constructed three AI centres in China and Israel as a collaboration between CASIA and the University of Haifa. In July, China, France and the Netherlands renewed an agreement for a joint Sino-European Laboratory in Computer Science, Automation and Applied Mathematics, in partnership with CASIA with a major focus on AI.
If there was one turning point in Chinese military attitudes towards AI it was in March 2016 during the World Go Summit when Google-owned DeepMind’s AlphaGo beat world champion, Lee Sedol. Lee’s defeat ‘captured the PLA’s imagination at the highest levels, sparking high-level seminars and symposiums on the topic’, the report said.
There was also a rise in Chinese military analysis of the US Defense Advanced Research Projects Agency’s programme Deep Green, which is a system that supports commanders’ decision-making on the battlefield through advanced predictive capabilities, including the ‘generation of courses of action, evaluation of options and assessment of the impact of decisions’.
As recently as September, the China Institute of Command and Control (CICC)sponsored the first Artificial Intelligence and War-Gaming National Finals, convened at the National Defense University’s Joint Operations College. ‘It involved a ‘human-machine confrontation’ between top teams and an AI system called CASIA-Prophet 1.0, which was victorious over human teams by a score of 7 to 1.’
Chinese military thinkers now want ‘intelligentisation of warfare that could result in a trend toward battlefield singularity’. Under these conditions, humans would no longer have capacity to remain directly ‘in the loop’, but would still possess ultimate decision-making authority or ‘human on the loop’, i.e. ‘exercising supervisory control’.
Chinese military strategists want to develop synergies between intelligentised or autonomous systems and directed-energy weapons that will enable ‘light warfare’ involving the fusion of real-time information and ‘zero-hour’ attacks. This will include all forms of military weapons. Chinese AI start-up IFlytek is working with the Chinese military on a voice recognition and synthesis module for intelligence processing for this very reason.
Of particular concern is the Chinese military’s Strategic Support Force (SSF) that seeks to build up advanced cyber warfare capabilities, leveraging big data and machine learning. According to the report, the SSF’s Information Engineering University has developed methods to detect and mitigate distributed denial of service (DDoS) attacks through pattern matching, statistical analysis and machine learning, as well as to detect advanced persistent threat detection based on big data analysis.
China’s national strategy of military-civil fusion enables China to transfer dual-use technological advances to build up military capabilities while promoting economic growth. The report advised the US government to compete and counter Chinese AI advances, and the Pentagon should consider supporting research to track the China’s AI defence innovation ecosystem.
Further, the report recommended reforms to laws designed to constrain ‘illicit and problematic’ technology transfers and changes on how the Committee on Foreign Investment decides what investments and acquisitions are a threat to national security.