AI in the War against Covid-19 in China:
An Integrative Coordination Governance (ICG)
March 2021
I. AI in the War against Covid-19 in China: An Integrative Coordination Governance
II. Manifestations of Integrative Coordination Governance
1. In Decision-making: An Integrative Design Mechanism for “Governance Products”
2. In Implementation: A Coordinated “Assembly Line” Governance by the Government and Private Sector
III. Two Indicators for Integrative Coordination Governance
1. Integration: Degree of Integration of Governance Motives and Capabilities
2. Infrastructure Maturity: Degree of Coverage of Hub Units
IV. Strengths of Integrative Coordination Governance
1. The Ability to Realize an Algorithmically-thinking Governance
2. Cross-issue Portability
3. Agility
V. Causes and Conditions for Integrative Coordination Governance
1. A Large Presence of AI Enterprises and NGOs with AI Teams and Infrastructure
2. An AI-adaptive Government
3. “Hub Units” within Governments, Enterprises and NGOs as Interfaces
VI. Conclusion: An Integrative Coordination Governance is Essential in the Era of AI
I. AI in the War against Covid-19 in China: An Integrative Coordination Governance
This report argues that the Covid-19 epidemic in China has been brought under control through an “integrative coordination governance” (ICG) achieved jointly by the government and society. Since the onset of the outbreak, emergency pandemic response and governance have been confronted with a host of highly complex issues. The intricacies of pandemic governance warrant the participation of artificial intelligence (AI) technologies; while the contingent nature of pandemic governance and the characteristics of AI technologies require that AI engagement in pandemic governance be delivered by a “new type of intelligent governance community” with a high degree of coupling between government departments, technology enterprises, and non-governmental organizations (NGOs).
This report analyzes, through case studies, the participation of different AI technologies in epidemic prevention and post-epidemic economic governance in China, and investigates the ICG mechanism that has empowered AI engagement in pandemic governance. The key findings are summarized as follows.
· The specific manifestations of ICG are: The decision-making body of the government remains in full control of the decision power; meanwhile, enterprises and NGOs are deeply engaged in facilitating government administrative affairs that need the support of AI technologies in the whole process by providing governance solutions before decision-making and implementation afterwards; in the whole governance process, the government and social sectors collaborate to form an assembly-line style of governance mechanism.
· There are two key performance indicators for ICG: integration and infrastructure maturity. The former refers to the degree of integration of governance motives and governance capabilities, and the latter the degree of coverage of hub units.
· ICG of the epidemic in China embodies three unique strengths: First, it enables an algorithmically-thinking governance; second, it enables cross-issue transference of governance capabilities; and third, it enables agile governance.
· Three conditions are required to deliver an ICG: First, a large presence of AI enterprises and NGOs equipped with AI teams and infrastructure; second, an AI-adaptive government; and third, effective interconnections between government, businesses, and NGOs.
This report argues that AI’s comprehensive and profound support for China’s epidemic governance is made possible through an ICG that fully mobilized enterprises and NGOs to apply algorithmic thinking for emergency public governance, thereby providing public services that are agile, targeted, and thorough.
AI engagement in China’s epidemic governance has pointed to an emerging and evolving ICG model led by an AI-adaptive government, which has inspired new perspectives and alternatives for AI engagement in governance that may be further discussed.
II. Manifestations of Integrative Coordination Governance
In ICG, the decision-making bodies of the government remain in full control of the decision power, while enterprises and NGOs are deeply engaged in facilitating government administrative affairs that need the support of AI technologies in the whole process by providing governance solutions before decision-making and implementation afterwards. It is worth emphasizing that in the governance process, the government and social sectors collaborate to form an assembly-line to complete various governance tasks.
1. In Decision-making: An Integrative Design Mechanism for “Governance Products”
At the pre-decision stage, businesses and NGOs play two roles: First, they tap into governance needs that require AI technologies and propose technical solutions. Compared to government departments, AI enterprises and NGOs are endowed with the following much-needed qualities for epidemic emergency governance.
· Well-supplied pools of AI technologies and practices
· Technology teams of great size and agility
· An innate ability to reduce complex governance problems to computational ones
These inherent competencies make companies and NGOs with AI capabilities indispensable in decoding epidemic governance demands and providing relevant solutions. In this way, they are empowered to translate their visions into the will of the government.
Case: A company taps into governance needs at the pre-decision stage
According to TMTpost, the design of the Health Code system was inspired by discussions of an Alipay team in their DingTalk group chat. Around January 28, a team member proposed in the group chat that governance efficiency could be significantly improved by switching from manual temperature measurement in subway stations to online reporting. The team started designing the product on that day, and on January 31, the Health Check-in system—the prototype of the Health Code system—was launched.
The public policy agenda proposed by the Alipay team made its way up the decision-making chain within the government. On February 4, the Yuhang District Government in Hangzhou motioned for a digital solution to epidemic prevention and control, and the Health Code project was immediately approved by the government and launched as the Health Code system on February 5. On February 6, a team at Hangzhou Municipal Public Security Bureau, including police officer Zhong Yi and other members, started to coordinate and optimize the design of the product. On February 11, the Hangzhou Health Code was launched, which was subsequently approved and promoted by the provincial government and the State Council of China. The case shows that AI enterprises, backed by solid technological resources, are able to identify scenarios where AI is needed in epidemic governance.
Sources: 1. 7 Days! How did China build its health code system this fast?
CCTV News, February 7, 2020.
http://m.news.cctv.com/2020/02/17/ARTIgiNx4V6AEUr0qhi3PeK8200217.shtml
2. Digital Intelligence Insight: The story behind health codes
WeChat Official Account of Alibaba Cloud Research Center, January 18, 2021.
https://mp.weixin.qq.com/s/4MsxWsMVzXGpsjo6O_NEaw
It is important to point out that at the decision-making stage, the government remains the only actor that controls the power and authority of decision-making. Whether the solutions proposed by enterprises and NGOs will be elevated as the will of public powers depends on the approval of governmental decision-making bodies, whose decisions shall be followed when companies design and implement algorithmic systems. The engagement of enterprises and NGOs in exploring needs and undertaking governance tasks reflects the integrative and coordinated nature of ICG.
2. In Implementation: A Coordinated “Assembly Line” Governance by the Government and Private Sector
At the implementation stage, the complexity and high frequency of epidemic governance tasks make it necessary to rely on enterprises and NGOs with digital infrastructure such as algorithmic technologies, user groups, and databases, etc. For example, tracking and identifying the sources of infection require tracking and analyzing the movements of hundreds of millions of people, which is beyond the capacity of manual epidemic response mechanisms. Therefore, it is essential to enlist the help of AI technologies.
Case: AI companies implement governance tasks at the post-decision stage
Platforms such as Weibo, TikTok and WeChat have clarified misinformation regarding the pandemic, removed false posts, and closed related accounts. Other platform companies such as Meituan and SF Express require their takeout riders and couriers to hold negative Covid-19 nucleic acid test results to prevent infections caused by the platforms’ business operations.
Source: Seminar on AI against Covid-19: Lessons and Governance, Institute for AI International Governance (I-AIIG), Tsinghua University, November 21, 2020.
It is important to note that in the ICG structure, the government and the private sector collaborate in an assembly line style of governance—performing different tasks from workstation to workstation seamlessly in the governance process. This assembly line collaboration between the government and the private sector is characterized by two main features.
First, task breakdown. A complete governance project is broken down into several governance segments on the assembly line, setting apart administrative tasks that require authorization (project approval, personnel mobilization, etc.) from technical ones of high computational complexity (resources deployment, capacity planning, etc.).
Second, tasking. Upon work breakdown, a rational assignment of the broken-down tasks is ensured based on the resource endowments of different entities. Administrative tasks are completed by government departments, and technical ones enterprises. Through a meticulous breakdown of the governance tasks, the degree of role misalignment between government and businesses is reduced, with each empowered to give full play to their expertise so that governance blind spots are minimized, and low-cost, scientific governance is achieved.
A large number of hub units between the government and enterprises are at play in breaking down and assigning governance tasks. Like conveyors on assembly lines, these hub units connect the unbundled governance segments together with each other. When a segment on the assembly line is completed, the hub units convey the governance information, resources and demands it contains onto the next segment swiftly, accurately and smoothly. In this way, the resistance and misalignment between governance segments are eased and governance efficiency improved.
Case: “Pneumonia (Covid-19) Patients Seeking Help,” one of Sina Weibo’s “super topics,” enables an assembly line style of ICG
According to www.guancha.cn, at the peak of the outbreak in February 2020, Weibo initiated a “super topic” on its platform titled “Pneumonia Patients Seeking Help” amid increasing user demand, providing a channel for Covid-19 patients to solicit assistance. The “super topic” community allowed Covid-19 patients and their families to leave their details which were collected in real time and handed off to the government by Weibo in collaboration with its media partners and volunteers. Concerned government departments, meanwhile, set up dedicated channels to verify such information with the help seekers. The information delivery service became a lifeline for the patients and their families.
Source: Pneumonia Patients Can Now Seek Help via Weibo Super Topic with Dedicated Government Channels, www.guancha.cn, February 5, 2020.
https://www.guancha.cn/ChanJing/2020_02_05_534647.shtml
III. Two Indicators of Integrative Coordination Governance
1. Integration: Degree of Integration of Governance Motives and Capabilities
In the AI-participating ICG, the boundary between governance motives and governance technologies is blurred or even disappearing. The government, with its governance power, strives for technology; enterprises and NGOs, with their AI technologies, have a motive for governance. In fact, an important characteristic of ICG is the inseparability of governance motives and governance capabilities.
A. Proactive corporate engagement in governance
Enterprises have demonstrated a strong sense of corporate social responsibility by proactively participating in and contributing to epidemic governance, which is manifested in the following two main aspects.
First, they proactively tapped into the demand for AI technologies in epidemic governance. In China’s epidemic governance, many technology companies and NGOs have taken the initiative to apply capabilities fostered in the business realm to perceptively identify AI task scenarios (i.e., to reduce complex problems to computational ones) in emergency public governance of the epidemic, and have succeeded in uncovering a string of governance scenarios where AI could play a part. For example, by analyzing its takeaway delivery services—a business scenario—Meituan caught on the necessity—the governance needs—to ensure a steady flow in its urban supply chain and to provide no-contact delivery during the outbreak. The company then used AI technologies to optimize its supply chain and delivery process. As a further example, ByteDance, a veteran in exploring business scenarios on its social media platform, quite observantly saw the societal risks brought by information asymmetry during the pandemic, and acted to optimize its recommendation algorithm and quickly launched several Covid-19 content projects to combat misinformation related to the pandemic.
Second, many enterprises and NGOs voluntarily bore part of the costs incurred in providing public services. In addition to out-of-pocket upfront payments, many enterprises and NGOs in China, during the ICG, chose to bear the costs of public service provision on their own. They would first propose AI governance scenarios and then invest in them voluntarily, not per governmental directives or requisition. It is noteworthy that the proactiveness of businesses and NGOs in ICG markedly sets it apart from traditional government procurement, public-private partnerships (PPP) or other reactive response models.
B. Algorithmic Thinking in Government Governance
As AI technologies were applied in epidemic governance in China, there have been cases where public governance problems were modeled and reduced to those of algorithmic design. Upon such modeling, efficient governance approaches for ICG can be identified by algorithms so as to reduce the burden of resource-strapped frontline governance workers, and AI algorithmic systems can be used to replace or collaborate with traditional administrative departments for more targeted and thorough governance.
Case: Algorithmic thinking in frontline epidemic response
During the June outbreak of Covid-19 in Beijing, health authorities applied a Divide-and-Conquer strategy against it. Specifically, each test kit was shared among a group of five, and if any of the tests came back positive, the five people using the kit in question would be re-tested separately. Since most people are not infected with the coronavirus, this strategy can augment the testing capacity by a factor of five in the same amount of time.
Sources: 1. What we talk about when we talk about algorithms: a divide-and-Conquer algorithm for Covid-19 nucleic acid testing, June 29, 2020.
https://v3u.cn/a_id_159
2. Collective procurement and price-haggling of test kits, ChinaTimes NetEase, June 24, 2020.
https://www.163.com/dy/article/FFU4F0770512D03F.html
2. Infrastructure Maturity: Degree of Coverage of Hub Units
It has been seen in many cases that ICG for the epidemic in China relies heavily on hub units—an organizational infrastructure that exists at all levels of government, businesses, and NGOs.
In fact, the hub units were already taking shape prior to the onset of the Covid-19 outbreak. In the pre-Covid China, AI technologies were already extensively involved in public governance in the areas of intelligent transportation and security, among others. These governance processes bred normalized, sectoralized and institutionalized interactions between government and the business world, leading to the establishment of many hub units connecting the government, enterprises and NGOs together.
In ICG, the hub units, along with supporting regulations, enable not only prompt design of emergency governance solutions, but also fast-tracked administrative approval procedures through the pipeline between government and enterprises, which is built on mutual trust, so that specific action strategies can materialize.
These hub units have also served as a seedbed for talents with cross-sectional skills in identifying public governance needs, developing governance solutions, and driving resource integration with computational thinking. In epidemic emergency governance, these talents have been the catalyst for the acceleration of algorithmic modeling of governance tasks and for the legislation of AI governance solutions.
The hubs have played an important role in bridging tasks, information and resources between government and businesses. Enterprises and NGOs who are ready to engage in governance would receive prompt responses from the government and by that, the latter’s emergency governance needs are translated into AI algorithmic tasks without delay. The hubs are arguably playing the role of “interface devices.”
Case: A government-business hub unit and its talents with cross-sectional skills
According to CZTV.COM, the Office of the CCP Committee in Yuhang District, Hangzhou, served as a hub unit for the rollout of the district’s health code system, known as Yuhang Green Code. The Office is reported to have functioned, during the development process, as the point of contact for technical teams at Alibaba. Indeed, the public officials at the Office have effectively turned themselves into “product managers.” Furthermore, the Office is the holder of the intellectual property rights of Yuhang Green Code, and Wang Junyi, its chief author, is a civil servant at the Office.
Source: 7 Days! How did China build its health code system this fast?
CCTV News, February 7, 2020.
http://m.news.cctv.com/2020/02/17/ARTIgiNx4V6AEUr0qhi3PeK8200217.shtml
IV. Strengths of Integrative Coordination Governance
1. The Ability to Realize an Algorithmically-thinking Governance
ICG has accentuated three unique strengths in China’s outbreak governance. First, it enables an algorithmically-thinking governance. As AI technologies were applied in epidemic governance in China, there have been many cases where public governance problems were modeled and reduced to those of algorithmic design. Upon such modeling, efficient governance approaches for ICG can be identified by algorithms so as to reduce the burden of resource-strapped frontline governance workers, and AI algorithmic systems can be used to replace or collaborate with traditional administrative departments for more targeted and thorough governance.
Case: Algorithmic thinking in frontline epidemic response
During the June outbreak of Covid-19 in Beijing, health authorities applied a Divide-and-Conquer strategy against it. Specifically, each test kit was shared among a group of five, and if any of the tests came back positive, the five people using the kit in question would be re-tested separately. Since the majority of people are not infected with the coronavirus, this strategy can augment the testing capacity by a factor of five in the same amount of time.
Sources: 1. What we talk about when we talk about algorithms: a divide-and-conquer algorithm for Covid-19 nucleic acid testing, June 29, 2020.
https://v3u.cn/a_id_159
2. Collective procurement and price-haggling of test kits, ChinaTimes NetEase, June 24, 2020.
https://www.163.com/dy/article/FFU4F0770512D03F.html
2. Cross-issue Portability
In ICG, the portability of AI technologies, together with the productive and organizational models of enterprises and NGOs, allow governance capabilities to be generalized and transferred to different issues.
During the process, the role of AI portability is specifically demonstrated by:
Ø Portable modeling capabilities. Portable modeling capabilities, which are technically reduction capabilities in computational science, have enabled public governance problems in other domains to be modeled into AI algorithmic tasks that can be transferred to epidemic public governance.
Case: Transference of modeling capabilities at the Institute of Public and Environmental Affairs
Before the pandemic, the Institute of Public and Environmental Affairs modeled the public supply of environmental information as a computational problem of data integration and visualization. The design was duplicated immediately, when Covid-19 broke out, to model the public supply of epidemic information as a parallel computational problem. By January 27, the NGO finished the integration of outbreak data and produced a map with visualized outbreak situation across districts and counties.
Ø Portable governance infrastructure. ICG in other public issues has employed and built a range of infrastructure in hardware, software, and virtual networks, including databases, user groups, and social networks, which can be put into service for epidemic emergency governance.
Case: Transference of social network infrastructure in epidemic response
The connections between TikTok and its millions of users has served as infrastructure for poverty alleviation and other public governance purposes before the pandemic, and when Covid struck, they were used by the company to build a dedicated channel to disseminate latest pandemic information to TikTok users. In addition, the databases for physical mobility tracking held by mobile network operators are also a kind of infrastructure which was embedded in the Health Kit system to facilitate epidemiological investigations.
Source: Yangtse Evening Post: Doctor went viral on TikTok for Covid-19 prevention video with 810 million views, eastday.com, February 23, 2020.
http://news.eastday.com/eastday/13news/auto/news/china/20200223/u7ai9112594.html
Ø Portable algorithmic systems (partially). In other public governance fields, various AI algorithms have been developed and employed for ICG. Since epidemic governance and other public governance issues can all be reduced to either identical or similar algorithmic problems in computational science, AI algorithms used for other governance issues can also be used directly, or upon modification, for epidemic ICG.
Case: Transference of face recognition algorithms
Before the pandemic, face recognition algorithms built upon visible light image analysis were the principal businesses of face recognition companies such as Megvii and SenseTime. When the outbreak started, these companies made adaptions to their algorithms for visible light image analysis and transferred them to infrared temperature measurement, which was then used to measure and record body temperatures in train and subway stations, bus terminals, airports, and other busy areas.
Source: SenseTime rolls out AI solutions with non-contact temperature measurement for epidemic prevention, SenseTime Baidu Official Account.
https://baijiahao.baidu.com/s?id= 1658500962900561346&wfr=spider&for=pc
The abovementioned threefold portability of AI technologies allows ICG developed in other public governance areas to be quickly adapted and used for epidemic governance.
Another reason why ICG in other areas can be promptly employed in epidemic governance is that the productive and organizational models of AI-capable companies and NGOs have the competency to transfer their capabilities across different issues. Amidst fierce competition in AI industries, new business potentials are constantly emerging. To stay competitive, companies must be able to swiftly translate their know-how from existing businesses into the capabilities to learn and develop products in new business territories. For example, just as community group-buying surfaced, AI companies soon jumped on the bandwagon by setting up teams and launching products without mass recruitment or team-rebuilding. Such organizational competency has prepared them to quickly apply their expertise in other governance domains to epidemic governance tasks.
Case: From livestream shopping to support farmers to “Mayors Show Hubei to You”
During the epidemic, a campaign called “Mayors Show Hubei to You” was launched by ByteDance, aimed to promote farm produce as an effort to support farmers in epidemic-stricken areas, who were provided with a new sales channel—a livestream platform—as well as selling skills training to sell their produce online. Before the outbreak, the project team had launched a similar livestream-shopping campaign meant to boost the livelihood of farmers. When the epidemic started, the team was empowered with increased human and financial resources so that a smooth and agile transition to emergency epidemic governance was actualized.
3. Agility
The agility of ICG is demonstrated by its ability to not only perceptively identify risks, but also to overcome challenges in data, technical, legal, financial and other dimensions in a very short period of time.
ICG is agile in uncovering the need for AI in epidemic governance: Within the ICG architecture in China, many technology companies and NGOs have taken the initiative to apply their capabilities fostered in the business realm to perceptively identify AI task scenarios (i.e., to reduce complex problems to computational ones) to emergency public governance of the epidemic, and have succeeded in pinning down a string of governance scenarios where AI could play a part. Therefore, governance needs and problems were identified without delay.
ICG can be an agile system that adapts to AI engagement in epidemic governance: It delivers not only swift design of emergency governance solutions, but also fast-tracked administrative approval procedures through the pipeline between government and enterprises, which is built on mutual trust, so that specific action strategies can materialize.
Through market-based resource scheduling mechanisms, ICG secures the feasibility of AI-designed governance solutions. Through market-based organizational and scheduling mechanisms within enterprises and NGOs, complemented by the administrative competence of the government, ICG ensures the feasibility of AI-designed governance solutions.
Case: The agility of ICG
On January 23, as the central government of China imposed a lockdown in Wuhan, governments and companies nationwide quickly developed IT products for epidemic governance.
· According to TMTpost, an Alipay team started out on the development of a health code prototype on January 28, which was launched on January 31. On February 5, the Health Code was launched, which was subsequently approved and promoted by the provincial government of Zhejiang and the State Council of China.
· On January 28, the Municipal CPC Committee and Government of Ganzhou, Jiangxi Province, proposed the establishment of an epidemic response platform. By January 31, the platform was up and running.
Source: Alibaba and Tencent entering the game: Can health codes replace passes? TMTpost, at tech.sina.com.cn, February 22, 2020.
https://tech.sina.com.cn/roll/2020-02-22/doc-iimxyqvz4878524.shtml
V. Causes and Conditions for Integrative Coordination Governance
Three conditions are required to realize an ICG: First, a large presence of AI enterprises and NGOs equipped with AI teams and infrastructure; second, an AI-adaptive government; and third, effective interconnections between government, businesses, and NGOs.
1. A Large Presence of AI Enterprises and NGOs with AI Teams and Infrastructure
Large AI firms have been proactive in governance because of their special nature—they need to install trust in emerging technologies during the legislation process concerning AI commercialization. Today, the legislation related to AI commercialization is evolving in a tug of war; technology companies are holding on to their newborn businesses which can be taken away any minute. As a result, technology giants have a strong incentive to gain the trust of the public and the government in order to further their business operations.
Specifically, AI companies seek to build trust for two reasons.
First, AI companies and NGOs rely on personal data for their primary businesses, which requires authorization for proper use of data. Thus, they need to foster public and government trust in them. As AI technologies are highly data-dependent, companies and NGOs that live on such technologies habitually need to collect personal data from the public, triggering ethical concerns regarding data privacy. Therefore, public approval for the access to personal data is a must for AI companies and NGOs that depend their survival on AI technologies.
Case: The face recognition controversy and proactive corporate response against Covid-19
Face recognition technology relies on the collection of biometric data about a person’s face and is therefore highly controversial, with many against its application. During the epidemic, face recognition companies such as Megvii and SenseTime took the initiative to develop temperature measurement technology which was later used to help identify transmission risks in train and subway stations, bus terminals, airports, and other busy areas, thus proving that the collection of biometric data about a person’s face and the use of face recognition technology can actually create value for society.
Second, principal businesses of AI giants feature a public character, hence the need for trust. As most of the main businesses of technology giants affect a massive population and thus indicate broad implications across the society, they carry, inevitably, a public character and will trigger controversies and regulations. For example, Shunfengche, the car-pooling service on Didi sparked a fierce debate and was eventually forced to a halt by the government after repeated incidents. Inspiring public confidence in their major businesses, therefore, becomes an essential goal for technology behemoths.
In summation, as a consequence of the above two reasons, AI companies and NGOs are driven to gain public and government trust. At the same time, the portability and generalizability of AI technologies make it less costly for companies to transfer these technologies from their central business to epidemic governance. Thus, such companies have shown a strong sense of social responsibility and initiative to engage in epidemic governance.
In other words, the urgency of trust-building and the portability of AI algorithms in their chief businesses are the ingredients for the companies’ motivation to fulfill their social responsibility. As set forth above, the proactive engagement of companies with AI technologies in governance is a key component of the integration of governance motives and capabilities, which is one of the key performance indicators for ICG.
Case: Food safety in the takeaway industry
Large takeaway platforms such as Meituan and Ele.me have always faced the challenge of food safety—and the legitimacy of their business is at stake. Worse still, the Covid-19 pandemic has raised fears over catching the virus through food. To address the issue, takeaway platforms have taken the initiative to employ AI technologies and design algorithms in order to prevent transmission of the coronavirus in their food supply chains and to protect deliver riders and customers during the outbreak.
Source: Meituan launches catering service guide against Covid with 4 industry bodies, China Economic Net, February 4, 2020.
http://www.ce.cn/cysc/sp/info/202002/04/t20200204_34214596.shtml
2. An AI-adaptive Government
China’s Covid-19 governance shows that the key to ICG is a government that understands AI technologies and has the experience and systems to accommodate AI companies in governance, or an “AI-adaptive government” as this report calls it. The main characteristics of an AI-adaptive government are as follows. First, it is motivated to improve governance effectiveness by envisioning scenarios with AI technologies. Second, it is experienced and equipped with frameworks to accommodate and regulate AI companies and NGOs in its governance system.
It is important to note that this motivation is a capability that has to be cultivated over time, and many cases show that it is more often found in regions where AI companies are alive and thriving, such as Beijing and Hangzhou.
On February 6, 2020, the Ministry of Industry and Information Technology of China released an initiative which calls on giving full play to the utility of AI technologies in combatting Covid-19. The initiative proposes to fully explore the application scenarios for the diagnosis and treatment of Covid-19 and for epidemic prevention and control. It also calls on striving for scientific and technological breakthroughs in the development and mass production of products for diagnosis support, rapid testing, intelligent equipment, accurate temperature measurement and target identification, in order to deliver intelligent Covid-19 diagnosis and treatment, reduce infectious risks for health care workers, and improve the efficiency of epidemic response. The initiative also calls for efforts to resume work and production by encouraging telecommuting, videoconferencing services and AI-empowered educational resources to facilitate remote working, distance education and intelligent production, so that work, schooling and production would not be disrupted during the epidemic. In addition, the initiative presses for optimizing AI algorithms and computing power to support virus genomic sequencing, vaccine/drug development, protein screening and other R&D breakthroughs.
Source: Giving full play to the utility of AI technologies against Covid-19, Ministry of Industry and Information Technology of China, February 4, 2020.
https://www.miit.gov.cn/ztzl/rdzt/xxgzbdgrdfyyqfkgz/gzdt/art/2020/art_f8660a2b7fcb44028ee8d7914e03f125.html
3. “Hub Units” within Governments, Enterprises and NGOs as Interfaces
It has been seen in a large number of cases that ICG of the epidemic in China relies heavily on hub units bridging the government, businesses, and NGOs. The hub units, along with supporting regulations, enable not only prompt design of emergency governance solutions, but also fast-tracked administrative approval procedures through the pipeline between government and enterprises, which is built on mutual trust, so that specific action strategies can materialize. These hub units have also served as a seedbed for talents with cross-sectional skills in identifying public governance needs, developing governance solutions, and driving resource integration with computational thinking. In epidemic emergency governance, these talents have been the catalyst for the acceleration of algorithmic modeling of governance tasks and for the legislation of AI governance solutions.
Epidemic governance is emergency governance. It requires, prior to an outbreak, the presence of large AI companies, an AI-adaptive government and hub units in order for them to play a role in ICG of the epidemic.
Case: A government-business hub unit and its talents with cross-sectional skills
According to CZTV.COM, the Office of the CCP Committee in Yuhang District, Hangzhou, served as a hub unit for the rollout of the district’s health code system, known as Yuhang Green Code. The Office is reported to have functioned, during the development process, as the point of contact for technical teams at Alibaba. Indeed, the public officials at the Office have effectively turned themselves into “product managers.” Furthermore, the Office is the holder of the intellectual property rights of Yuhang Green Code, and Wang Junyi, its chief author, is a civil servant at the Office.
Source: 7 Days! Programmers in Hangzhou behind China’s health code, CZTV.Com.
http://i.cztv.com/view/13412619.html
VI. Conclusion: An Integrative Coordination Governance is Essential in the Era of AI
Covid-19 governance in countries show that governance in the era of AI faces three universal challenges.
First, though AI technologies are urgently needed for public governance, governments alone are not capable to use them for governance. Through the eye of AI algorithms, there is a host of complex, high-frequency problems in economic and social governance where AI technologies are undeniably needed. Particularly for emergency governance such as epidemic prevention, control and recovery, AI technologies are able to facilitate accurate, targeted and thorough design and execution of governance solutions, which, even more explicitly, points to the urgency of AI engagement in public governance. However, AI technologies are highly dependent on large teams of computational engineers and high-cost computing resources, which means that modern bureaucratic governments are organizationally incapable of employing and developing such technologies.
Second, for companies and NGOs equipped with AI technologies, as their central businesses are data-dependent and thus have a public character, they would fatedly seek public and government recognition in pursuit of political utility. In practice, they have already taken up a large number of public governance tasks in cyberspace. However, as part of the private sector, it would be hard for them to acquire legal authorization for their engagement in public governance.
Third, traditional public service procurement or PPP cannot support the engagement of AI technologies, a highly challenging discipline, in public governance. When AI technologies are used to explore public service demands, it relies on the know-how of the governance subject to model governance problems into those of AI algorithm design, and the development and implementation of AI governance solutions rely on strong computing power. This is all beyond the capability of modern bureaucratic governments, hence their inability to identify public service demands or to adopt old models such as conventional public service procurement or PPP.
AI engagement in China’s epidemic governance shows that AI technologies have ample potential to achieve proactive, agile, accurate, targeted and thorough governance, the effectiveness of which is built on its interactive structure along with its technical and organizational characteristics. First, in ICG delivered by government departments in collaboration with technology companies and NGOs, the latter two are proactively engaged in epidemic governance because they pursue the trust of the public and the government. Therefore, they endeavored to organize resources in a market-oriented manner and offered their AI capabilities. Second, the portability and generalizability of AI technologies allow the governance capabilities developed around one issue to be transferred to others swiftly, smoothly, and at low cost. Third, the resilience, perceptiveness, and target-oriented nature of talented AI teams provide robust organizational support for AI engagement in governance.
With rising public expectations of governments’ governance performance in the era of AI, all countries face the question of finding better ways of using AI technologies to strengthen their ability of addressing complex issues, thereby boosting governance performance and, ultimately, overall national competitiveness.
This study points out that AI engagement in China’s epidemic governance shows that the evolving economic and societal foundations in the era of AI require certain adjustments within bureaucratic governments in order for them to adapt to such transformations. ICG by the government and society together, as an emerging and evolving model, provides new perspectives and alternatives that may be further discussed for the design of governance structures in the era of AI.
