Is Automation Making Chinese Society More Resilient?

A look at China’s response to the COVID-19 pandemic through case studies of the automated systems enabling urban and socio-political resilience.

In combating the COVID-19 pandemic quick actions and rational decisions are extremely important. However, many countries, including Western ones, have failed to effectively contain the crisis. Perhaps one reason could be that they are limited within biopolitics in which automated technologies, such as sensing and indexing, are categorized as negative surveillance. This poses an obstacle to the effective use of technologies, according to Benjamin Bratton. He argues that there should be a resilient shift in the biopolitics of governance—from humans to the management of ecosystems directly—with the principle of automation.

From Bratton’s point of view, automation is not simply technology, but more broadly a principle on which machines, humans, minerals, and methods rearrange. This principle means that decisions are made and actions are taken by technical systems which sense, index, and calculate geological facts, rather than by self-conscious humans. With this principle of rearrangement, the systems can become more resilient, with resilience referring to a person and a system’s capacity to adapt to changes and recover. Unlike the resilience of elastic materials, which is simply in their nature, the resilience of systems is developed by artificial efforts, including observing, making decisions, and taking technological actions. In this sense, pervasive automation and system resilience can converge.

This essay will examine whether and how Chinese society became more resilient by using various technologies or by following the principle of automation during the COVID-19 crisis. These automated technologies mainly operate in two ways: top-down enforcement and bottom-up feedback. Top-down enforcement includes systematic plans, algorithmic decisions, and governance via infrastructure and algorithms for damage prevention and mitigation. The bottom-up approach includes collective creative engagement through digital networks, aimed at overcoming the government’s opacity. Various automated technologies work mainly in these two ways, transforming Chinese society’s structures and politics.

Sensing and Prevention

The first case deals with top-down enforcement. When a crisis or emergency strikes, it is important to systematically understand the situation as it changes. Systematic sensing and measuring are crucial for the acquisition of data for better responses. Sensing inevitably raises privacy or surveillance concerns. However, the objective of the observation here shifts from human behavior and desire to the physical substrate, including the material transformation of biochemistry, regional ecologies, cities, and even the virus.

Various sensors have been extended from industrial areas into daily life, such as smart devices, transport, satellites, and so on. They turn the urban setting into an automation system or “machine landscape”—as Liam Young describes it—and sense, model, and calculate it. At times, sensing might be broken. Bratton argues that the sensing layer was broken in some countries during the COVID-19 crisis because of insufficient testing and inadequate preparation and governance.

In China, despite insufficient testing initially, the existing information systems and infrastructure tracked and visualized the pandemic risk situation. Prior to Wuhan’s lockdown, about five million migrants moved around the country to celebrate the Chinese New Year holiday in other cities. Those population flows can be tracked as virus-spreading movements at the early stage in China, mainly due to the coordination of different sensing infrastructures, such as the two types of data systems explained below.

One type is AI-powered mapping systems, including the Baidu Migration platform and Tencent Location platform. They traced and recorded population movements, especially to determine where and when they arrived. For example, Baidu Migration showed that 74.77 percent of the population it was tracking left Wuhan for adjacent cities of Hubei Province on January 22, 2020. Such data enhanced recipient cities’ prediction abilities and preparation—such as mobilizing doctors and medical equipment—and increased their response efforts. In fact, such migration data was subsequently shown to be highly predictive.

The WorldPop research group at Southampton University found that there is a linear correlation between the population leaving Wuhan to destination cities and COVID-19 cases in these destinations at an early stage. They analyzed the correlation between the number of imported cases reported and the risk of travelers from Wuhan importing the virus within the two weeks before Lunar New Year’s Day.

Another Chinese study adopts a similar method, based on a wider range of migration data sources from both Baidu and Tencent, and more importantly considers an incubation period of three to seven days. It calculates the correlation between the “imported risk” at the city level and the total number of cases in that city. This study concluded that along with time, the correlation between these two variables tended to level off at 0.92 around January 31, 2020.

Both studies demonstrate that at the early stage, most of the cases were related to travelers from Wuhan. In this sense, actively tracing the transmission chains of the confirmed cases on various means of transport played an important role in effective control and prevention of the pandemic in China.

Another type of data collection system is the real-name ticketing system, which helps narrow down potentially infected persons and virus carriers. In China, all passengers must use their state-issued identification card to buy public transport tickets and ride on public transport, and the state ID database has the names, addresses, and other information of all Chinese citizens. Once passengers who took trains had been confirmed as having COVID-19, the system could quickly identify those who took the same trains, since the ticketing system records passengers’ train and seat numbers, as well as their IDs. Authorities could ask those who had contact with the infected people to self-quarantine or place them under medical observation.

Here, the machine vision from these two systems mapped and visualized information, facilitating a better understanding of emerging trends in real-time. With sufficient and continuous information from the infrastructure, local governments and public health departments were able to make a comprehensive plan to alleviate the damage and help prevent further contagion.

Algorithmic Governance: Health QR Code

The second case also deals with top-down enforcement. Based on a robust sensing infrastructure, traditional politics in China are quickly shifting to algorithmic governance. One prominent aspect of this shift is the transformation of the decision-making process, underpinned by infrastructure and geotechnologies. For Bratton, in contrast to self-conscious decision-making, geotechnologies can automatically generate decisions repetitively, since geotechnologies can measure, simulate, and calculate the material substrate and receive feedback from users. The human role during the decision-making process could at times give way to an infrastructure-based automation.

In China, an example of this transformation is the health QR codes which were rapidly developed during the COVID-19 pandemic. The decision of who should be regarded as infected or potentially infected and to be quarantined is difficult to answer by traditional health staff. However, the answers are already embedded in infrastructure and other technical systems. Using data about an individual’s travel and medical treatment histories and social contacts, sensed by various geotechnologies and algorithms in the health QR code system, automatically and quickly evaluate users’ risk levels, summarizing with an easily understood “traffic light” result.

With the rise of automatic decision-making, the established traditional government is gradually losing ground, some of it being occupied by “platform sovereignty,” to put it in Bratton’s terms. For example, the above-mentioned health QR codes were rapidly developed by Chinese private-sector tech companies Alibaba and Tencent in February 2020 without the central government’s oversight. Local governments quickly adopted them and then relied largely upon them. In almost every city, before entering public places, everyone has to show their health QR code: green allows access, while yellow and red are denied access. Red is also reported to the authorities.

In order to harness the strengths of tech-giants, the Chinese central government enacted regulations in May 2020 to standardize the health code’s reference model, data format, and application interface. The national standardization of health codes ensures that all cities adopt the same set of algorithms and data sources to increase the accuracy of the evaluations and the equality of automated decisions by avoiding inconsistencies in different areas of the country. It also allows individuals to travel within China without encountering unexpected difficulties.

Despite these regulations, the pandemic has accelerated the Chinese government’s efforts to transform into a platform sovereignty that embeds automated decisions. At the end of May 2020, the city of Hangzhou Health Commission shared the idea of the gradient health QR code. The system will collect multiple layers of data, including the user’s digital medical records, lifestyle, and so on, and evaluate them with the same value unit, the point. For example, by drinking 200ml of alcohol the user will lose 1.5 points. Those points will be synthesized as a precise measurement to depict users’ health conditions, displayed on the interface as a gradient color range. In this way, health management might become more convenient, with detailed causes and directions.

Algorithmic governance through digital platforms was practiced in China for effective governance even before the pandemic. Alibaba had already established the City Brain system for some local governments. The City Brain system collects, processes, and combines various levels of data—such as transport, community and public safety, and city events—and displays them on the same large interface in the governor’s office which also shows data referring to transport, the public health system, public services, and so on. The Chinese tech giant Tencent is also developing “City QR code”, which aims to establish a daily digital public service platform. These platforms are improving city governance as well as gradually absorbing governmental functions, as many plans and decisions are quite dependent on these platforms.

A further direction of these systems should emphasize the prevention of future crises—such as pandemics or ecological disasters—by managing current and potential risks, which would make them more resilient. A typical example, although not about this pandemic, is the Emergency Command System of the State Grid Corporation of China, which monitors electricity networks in China and can automatically flag emergencies and update situations in real-time.

From the aforementioned cases, we can see that automation as a principle or logic has gradually permeated into Chinese society governance. Infrastructure, robust agency, algorithms, and humans have merged into a system.

Contributory Work and Techno-diversity

Indeed, by using technologies for better observation and wiser enforcement, a stronger government could become more resilient, but there are also limitations of the opaque top-down process. One approach to alleviate or supplement is the bottom-up approach of providing more feedback. That approach is reflected in the contributory discussion and creative engagement of Chinese individuals based on digitalization, which could also lead to what philosopher Yuk Hui calls “techno-diversity,” which in turn will “produce new thinking that integrates modern technology into their traditions and also transforms those traditions”. This new thinking can help discover new ways of combating pandemics in the present moment, and in the longer term develop new lifestyles, new principles of economics, and new forms of decarbonization technologies, as well as help reduce consumption. In other words, achieving techno-diversity is to think and create an alternative world, different from the anthropocentric one where only certain cultures dominate the world.

During the COVID-19 crisis, many Chinese citizens have creatively dealt with authorities’ opacity by using GitHub and WeChat. GitHub is the world’s largest community of open-source software, while WeChat is the country’s most popular messaging, social media and mobile payment app. What did Chinese people gain from using GitHub during the pandemic, and for what reasons?

First, they used GitHub to circumvent authorities’ opacity and censorship. GitHub provides free storage space for public and private repositories, and due to its decentralized and distributed structure, everyone who has contributed to a GitHub project has a copy on their own computers. In contrast, the centralized service of established Chinese online networks can be controlled and information easily deleted. Taking advantage of the decentralized feature, ordinary Chinese citizens used GitHub to help maintain transparency during the COVID-19 pandemic. This helped give real feedback to the governance systems for them to improve. They created repositories such as 2019nCovmemory to store information, which were banned or deleted, especially those about the Wuhan local government’s early-stage reactions which failed to warn the public of the novel coronavirus and facilitate testing.

For example, in the early stages of the COVID-19 crisis in China, an article featuring a first-person account by Ai Fen, a doctor who shared the first report of the novel coronavirus with her colleagues and her university friends who work at other hospitals, was censored by the Chinese government. While doctor Li Wengliang—who later died of COVID-19—was praised as the whistleblower, Ai Fen was nicknamed the “whistle-giver.” Such articles help communicate the true situation and show how a society reacts in the midst of a crisis. Making a crisis traceable is helpful for effective crisis management, and benefits resilience, while opacity and censorship serve as obstacles to early detection and the ability to react quickly.

Second, GitHub supports contributory and collaborative works. It supports source code management, and users can adapt computer code written by others and modify it for their own purposes. For example, Tencent developed a COVID-19 self-triage system shared on GitHub. They encouraged the public to reuse their codes by modifying the questions and answers, which are used by pathological consultants, and to judge risk according to specific diagnostic guidance. Several resource management and distribution networks also emerged. For example, the GitHub repository “wuhan2020” collected reliable information about available resources and other services from hospitals, factories, and so on, to fight against the pandemic. Their open-source sub-projects were employed to automatically process information without manual work, and increased the efficiency of other data processing works. Indeed, this project has been very popular, as it has been reused 927 times.

Further development of openly shared information created engagement and solidarity amongst ordinary Chinese citizens. An impressive example is the re-invention of “the Whistle-Giver” report. Soon after it was deleted, many Chinese translated it into various languages, including English, ancient Chinese, Morse code, J.R Tolkien’s Elvish language, Star Trek’s Klingon language, ancient Hebrew, and so on, and shared it on WeChat. Since the information was posted in various languages, the posts were beyond the capability of filtration censorship mechanisms. Those creations and subsequent sharing of those creations also transformed GitHub and WeChat from a hosting service and social network to socio-political agents, respectively.

These technologies enabled transparency and re-shaped relationships amongst individuals. These various translated versions echoed and connected people with each other, and also gave feedback on governance and politics. An “us” emerged which represented a new collective willing to not only publicly expose the feedback of systemic problems, but also to work on transforming them. In this case, alternative and novel methods were invented by the collective, through the bottom-up approach, to overcome top-down opacity and censorship. This increased society’s ability to mitigate damage.

Towards more resilient governance

Chinese society underwent a wide range of experiments about the alternative use of surveillance technologies. These empirical examples provide some evidence that automated technologies could enable resilience through both top-down and bottom-up approaches. To achieve this, first and foremost is the idea of resilience in terms of the perspective towards technologies. Technologies never only belong to a certain aim or schema. Changing their connotation and developing their alternative functions are kinds of resilience. This asks for detaching functions such as power apparatus from automated technologies and reattaching other goals, such as more rational and sustainable governance.

Moreover, it is shown that Chinese society is accelerating the construction of machine ecology, in a way very close to the principle of automation. In the top-down approach, existing surveillance technologies in China are used for observing infrastructural situations in order to better understand the spread of the pandemic. The sensed information is then used for automatic decisions to effectively organize society while traditional governance methods are much slower and less efficient. They both facilitate top-down enforcement to be more rational, precise, and quicker. Facing the limitation of overenforcement, such as information opacity and censorship, the bottom-up approach provides a suitable supplement. Digital networks are used collectively and creatively by ordinary citizens to bring together incongruous information into the system, to make the system more adaptable and resilient. They both contribute to a system’s structure and abilities to face unexpected changes.

Speculating about the future, could the principles, if not exact methods, used in China be broadly applied to other countries? The insufficient adoption of automation principles in many countries leaves them less resilient in the face of crises like COVID-19. Governments and societies should explore and reinvent the role of technologies to use them towards building more resilient socio-political systems.

Cover image: Interface of City Brain in Xiaoshan district, Hangzhou City. Image courtesy of Alibaba

​

Jie Shen

Jie Shen is a postgraduate researcher in Western philosophy and The Terraforming remote fellow. Her research interests include technology of philosophy, automation and society, and the technological constitution of mind and desire.