AI Adoption by PLA, AI in Decision Making, World's First Twin-seater Stealth Fighter, Zuchongzhi 2, Classification of Internet Platforms, and CAC Data Regulations
Issue 5, 1 November 2021
Note: I generally publish the newsletter on weekends. However, this weekend turned out to be quite eventful for China tech watchers. CAC and SAMR released some important draft regulations (covered in Section II) and the Georgetown University’s CSET released an important report on PLA’s AI adoption in 2020 (covered in Section I). So I took some time to go through the reports and guidelines and I have highlighted key points from these documents in today’s newsletter. Enjoy!
I. Military and Warfare
Artificial Intelligence (AI) Assisted Decision Making
Authors Zhang Yuan and Tong Xinbo (affiliation not given) write in PLA Daily about how AI will help the decision making in the battlefield and "how to fight" in intelligent warfare (智能战争). Four ways AI could help future military leaders in decision making.
Improve awareness - AI can build a battlefield knowledge center that will aid commanders to convert information into knowledge. AI can be used to conduct in-depth research on command behavior, war event deduction, and analyzing historical cases, and form a knowledge map using big data, multi-source data fusion, and dynamic correlation. This battlefield knowledge center can act as a second brain for the commander.
Quickly generate solutions - AI can be used to match current mission scenarios with historical cases and find similarities and differences between the two. This can be used to build better and implementable tactical plans.
Battlefield game problem - AI can simulate the intelligent game between the red and blue parties, for combat training exercises. Using deep learning, reinforcement learning, and other methods, AI can generate and optimize strategic plans for games and exercises. The AI can optimize the battle plan in real-time, quickly recognize the subtle connections and influences of the current battlefield combat activities, and dynamically complete the high-target value threat estimation.
Deducting solutions - AI can analyze the efficiency of the deduced plan and give an estimate of victory or defeat, evaluation of the feasibility and effectiveness of the plan, point out the existing problems, and give optimization suggestions.
Information Optical Technology (信息光学技术)
Ever wondered what if you could see through walls? Well, there is a radar which can! And it is being used in various military applications.
A perspective radar that can see through walls can emit high-frequency electromagnetic waves with strong penetrating power to scan walls. The specially built detector receives the diffusely reflected echo of the signal and uses an algorithm to determine the situation behind the wall. The article mentions XAVER400 Imaging System developed by Israel which can see-through walls.
An article by researchers from the Academy of Military Sciences explains the Information Optical Technology (信息光学技术) and is being developed and used by some countries for military applications. The current military applications include synthetic aperture radar, through-wall perspective radar, and holographic imaging technology. The Synthetic Aperture Radar is a high-resolution radar that is used to scan objects to obtain two-dimensional or three-dimensional information. This is especially useful under extremely low visibility weather conditions and still gives high-resolution images of the scanned objects.
Holographic imaging technology to create a 3D battlefield in real-time. Will give real-time battlefield situation information and aid in decision-making for commanders. The article argues that such technology can also be used to develop illusion weapons, causing confusion, destroying the enemy's fighting will, visual camouflage of weapons and equipment, and create virtual false targets, etc.
Combat Formation (兵力编组)
An article in PLA Daily talks about the importance of combat formation, also known as "force formation" or "strength formation" (兵力编组) and its importance in combat planning. The authors detail 4 considerations for designing an impactful future combat formation. One such consideration is for intelligent fusion grouping (智能融合编组) which includes robot soldiers, unmanned units, intelligent assistants, specialized teams as a part of the combat force. The authors argue that human-machine mixed intelligence is an essential part of intelligent warfare.
World's First Twin-seater Stealth Fighter
Minnie Chan from SCMP reported that China is preparing the world's first Twin-seater stealth fighter, a version of the J-20 in a fortnight. The article speculates that this Twin-seater version of the "J-20 Weilong" (translated as J-20 Mighty Dragon) could make its debut as part of PLA Air Force's 72nd-anniversary celebrations. Some believe this is an attempt to prove that the US concept of next-generation air dominance (NGAD) could be successfully applied in Chinese aircraft technology.
The aircraft appears to be developed by the Chengdu Aircraft Industry Group (CAIG). Earlier in January this year, the state-owned Aviation Industry Corporation of China (AVIC) for the first time released a computer-generated image (CGI) of the twin-seat J-20 'fifth-generation multirole fighter aircraft in a promotional video. That time, it was not sure if China had developed such aircraft or is at a conceptual stage. However, latest developments show such aircraft has been developed. As reported by Janes, a promotional recruitment video released in this January for People Liberational Army Air Force (PLAAF) also indicated that J-20 fighters fitted with locally produced Shenyang-Liming WS10 'Taihang'-series turbofan engines have entered service with a training unit under the Central Theatre Command.
Global Times also reported that the images of this twin-seater stealth fighter were circulating on social media in China and went viral. The Global Times article states that the second pilot will allow the possibility of dealing with complex tasks that can't be done by one person alone. It can be used for electronic warfare, control accompanying drones, tactical bombing tasks, convert J-20 into the aerial command center and control a swarm of "loyal wingman-style drones", etc.
The Loyal Wingman is an Airpower Teaming System (ATS) developed by Boeing. It uses AI to perform autonomous support operations while flying alongside a manned aircraft. China Aerospace Science Technology Corp's (CASTC) also unveiled its "loyal wingman" style FH-97 drone concept at the Zhuhai Air show this year in September. The United Kingdom (UK), Australia, Russia, and India are also developing such drones.
How the Chinese Military is Adopting Artificial Intelligence
Centre for Security and Emerging Technology (CEST) released a report "Harnessed Lightning - How the Chinese Military is Adopting Artificial Intelligence" written by Ryan Fedasiuk, Jennifer Melot, and Ben Murphy. This report examines AI-related equipment contracts awarded by the PLA and China's state-owned defense enterprises in 2020 to assess how the Chinese military is adopting AI.
Key findings of this report are:
Chinese military leaders are already procuring AI-related systems and equipment to prepare for "intelligentized" warfare, but AI so far represents a small fraction of overall purchasing activity.
Chinese leaders view AI as the key to transforming the PLA into a "world-class," globally competitive military force. PLA advances in AI and autonomy will create new vulnerabilities for the United States and allied forces operating in the Indo-Pacific.
China's military-civil fusion (军民融合) development strategy is helping the PLA acquire Commercial off-the-shelf technologies, both from private Chinese technology companies and sources outside of China.
The report identifies seven primary application areas for which the Chinese military is awarding AI-related equipment contracts:
Autonomous vehicles
Intelligence analysis
Information warfare
Logistics
Training
Command and control
Target recognition
There is also a profile of 273 known AI suppliers of the PLA.
PLA Strategic Support Force (PLASSF), being a service branch for cyber and electronic warfare, turned out to be the most active in procuring AI-related technologies. PLA Ground Force (PLAGF) came second.
Moreover, the report concludes that of the 273 AI equipment suppliers in the dataset studied for this report, 60% are private companies. The majority of equipment suppliers are quite small and were established only in the last 10 yrs. Less than 10% of these suppliers face US export controls. Rather, some companies supplying AI to PLASSF are funded by US venture capital firms.
One of the authors of the report, Ryan Fedasiuk tweeted further observations that didn't make the cut in the final report.
Chinese military may have an edge over the US on artificial intelligence research, report warns SCMP coverage of the report by Linda Lew.
II. Non-military Updates
World's Fastest Quantum Computer - Zuchongzhi 2
China's quantum computing research has taken a leap in the last few years. After launching the world's first quantum satellite in 2016, Chinese researchers have now announced that they have built the world's fastest quantum supercomputer -Zuchongzhi 2. It is a 66-qubit programmable superconducting quantum computer that is claimed to be is 10 million times faster and more powerful than Google's 55-qubit Sycamore. The lead researcher of the project Pan Jianwei (潘建伟) and the team have published their findings in two papers published in the peer-reviewed academic journals Physical Review Letters and Science Bulletin. Jiuzhang 2, another quantum computer is based on light.
Classifying Internet Platforms
State Administration for Market Regulation (SAMR) will classify the internet platforms in China into super, large, medium, and small, based on their user scale, business type, and capacity. SAMR released two draft guidelines on October 29.
"Guidelines for Classification and Classification of Internet Platforms (Draft for Comments)"
"Guidelines for the Implementation of Subject Responsibilities of Internet Platforms (Draft for Comments)"
Guidelines for Classification and Classification of Internet Platforms (Draft for Comments)
The guidelines also define 6 types of internet platforms.
Guidelines for the Implementation of Subject Responsibilities of Internet Platforms (Draft for Comments)
This document outlines the responsibilities of internet platforms, especially the super larger platforms which have an advantage over larger, medium, and small platforms. Article 1 to 9 lay down the guidelines on fair competition, equal governance, promoting interoperability, data management and security, internal governance, risk assessment, risk prevention and control, security audit, and promoting innovation.
Unsurprisingly, Article 34 of the "Guidelines for the Implementation of Subject Responsibilities of Internet Platforms (Draft for Comments)" mentions that internet platform operators shall "actively cooperate" with the supervision and law enforcement agencies in performing their duties and "provide the law enforcement agencies with the data" and cooperate with the law enforcement agencies in "monitoring."
This is in continuation in the series of regulations being imposed on tech giants in China. Apart from the fines for monopolistic practices, China is also reportedly considering internet platforms to make third-party search results on their website and tear down the "digital walls" in China's internet platforms. For example, if you search, say, something like "Li Zhiqi" on Baidu (Google-like search engine in China), the results page won't display entries from WeChat (owned by Tencent) or Douyin (owned by ByteDance). China's Ministry of Industry and Information Technology (MIIT) along with regulatory agencies like Cyberspace Administration of China (CAC) are taking steps to address issues like monopolistic practices, security and privacy issues of China's tech giants.
CAC Releases Draft Regulation on Outbound Data
Cyberspace Administration of China (CAC) has released draft regulations for outbound data to safeguard national security and social interest under the "Network Security Law", "Data Security Law ", and "Personal Information Protection Law", etc.
Companies with more than 1 million Chinese users will have to do a security review before they can transfer any Chinese data abroad. This applied to all the Chinese companies and foreign companies operating in China, CAC will be the agency responsible for security reviews. The security review will be required if:
Data provided is personal information and important data collected and generated by operators of critical information infrastructure;
Contains important data;
Personal information processors who have processed the personal information of one million people provide personal information abroad;
Cumulatively providing personal information of more than 100,000 people or sensitive personal information of more than 10,000 people abroad;
Other situations required by the State Cyberspace Administration of China require data exit security assessments.
The criteria for needing security assessment are vague. For example, the "critical information infrastructure" and "important" data. Anything can be designated as important. Although, another draft regulation published by MIIT in September classifies data into three categories and bans the export of "core" data. The "core" and "important" data referred to in the CAC draft could be similar.
III. In Short
PLA conducted 1st underwater explosion destruction test in which a wharf simulated as hostile was destroyed. The article in Global Times argues that this tactic can be used to counter the US Navy.
China's 3rd aircraft carrier is expected to be launched in 2021.
Bloomberg reported Gregory Hayes, CEO of the US defense contractor Raytheon, has said the United States (US) is years behind China in its pursuit of hypersonic weapons.
The Ministry of Industry and Information Technology (MIIT) announced the list of pilot demonstration projects for the development of the big data industry in 2021.
The State Council issued the "Fourteenth Five-Year Plan for National Intellectual Property Protection and Utilization Plan."
MIIT released "Guidelines for the Construction of the Basic Security Standard System of the Internet of Things (2021 Edition)"
IV. Additional Readings
The Fall and Rise of Techno-Globalism, By Graham Webster and Justin Sherman, Foreign Affairs
"Baumol Effect" and Overcoming it Using Digital Technology, Pekingnology
ITOW: PLA Terms and Concepts: Training, Exercises, & Education, China Aerospace Studies Institute
I highly recommend this report by China Aerospace Studies Institute to anyone interested in PLA and is studying Mandarin. The report divides selected terms into categories of Training, Plans, Exercises, Education and Training, and Three Levels of PLAAF Cultivation and Training. The words seem difficult and impossible to remember at first, but understanding the meaning of these key terms would go a long way in understanding the PLA way of thinking.
Could you please write on PLA Strategic Support Force (SSF) which has been elusive for many China Watcher.