Please enter keywords
Technology Platforms in Harmony with Society
Date:09.07.2021 Author:LIU Xiaochun - Vice President, Shanghai Finance Institute (SFI)

Abstract: To make technology platform companies live in harmony with society, the concept and logic of data transactions should be clarified, and regulations should be released before the opening of data exchange markets. This article provides five policy suggestions: 1) regulating on the means of realizing market concentration rather than the fact of it; 2) simplifying platform functions; 3) a licensing system for data collection, governance, service, and sale, 4) establishing a separate supervision system for technology platform companies and the data industry; and 5) technology platforms must stay open.


For millions of years, mankind has pondered eternal topics such as the distinction between humans and other animals. They are asking themselves because no other species will come and answer.

The activities of a dozen elephants in Yunnan Province have recently piqued the interest of people all around the world. They go in formation by day and night, with those who lead and those who follow; they sleep, rest, and stand guard in formation; they play and fight with one another, and when they dispute, they go their separate ways. It may be deduced from the behaviors that they have language, feelings, and division of labor, as well as thoughts, and that they are social animals.

Therefore, it appears that the true distinction between humans and other creatures is in writing and science.

Writing and science are intertwined. Science has advanced human society and liberated humanity from complete reliance on natural resources. Humans are living longer lives with increasingly ample material resources. As a result, humanity reveres science, believing that it will bring them prosperity, progress, and happiness.

Throughout the centuries, science and technology have progressed and human life has evolved, but we still have to manage with aging, sickness, and death; human communities continue to be plagued by inequity and resource wars. Science and technology may have solved many of the survival issues, but they can't eliminate the problems, and in some cases, they even exacerbate them.

There was a time when people believed that science and technology would release mankind from heavy and repetitive manual labor, and society would be more equitable and peaceful. Such a better society, however, has never been realized. Tiresome physical labor persists albeit evolved with technological progress and has been pressed more accurately and brutally by capitals and technology.

Industrialization and the replacement of man by the machine came at the same time as Taylorism was invented. Machines, equipment, and production management systems lead to worker destitution and strained labor relations.

The starting point of the Internet, big data, cloud computing, and AI is to benefit humanity, but the logic of capital is to reduce labor and labor costs. The even more accurate and scientific management brings "996" working hour system and confines the employees, deliverymen, and customers through algorithms.

A while ago I spoke with a cook of Suzhou pastries, intangible cultural heritage, and learned that many famous cuisines came from leftover food ingredients from main dishes. To avoid food waste, the cooks cut and chopped them into small pastries and accidentally produced famous dishes and more profits.

In Hong Kong, I learned that some supermarkets and shopping malls give away the food nearing its expiration to the charitable groups they partner with. These food items, according to the rules, are not allowed to be restocked the next day. Of course, the stores will continue to summarize the proper amount of selling, stocking, and allocating, to limit the waste and save money.

However, big data, cloud computing, and AI firms don't research such laws as to commodity sales or serve to compute the number of goods and stocking scientifically and accurately. Excess goods are simply tossed away. To keep sufficient flows of customers, the malls also have to lower the prices. As a result, the only way left to keep costs down is to squeeze their upstream suppliers, making life worse for upstream suppliers and even farmers in the planting and farming industry.

To attract consumers, several companies that are listed at high valuations disregarded financial risks and over-loaned to low-income people, resulting in major social consequences. Such a business model is bound to wear itself away.

There was a time when we believed that with the development of the Internet, data would become more open and information more transparent; it'd be easier for sellers to find demanders and consumers to find the goods they need; investors could set aside intermediaries, directly locating suitable investment projects and assessing risks accurately. With this belief, all sectors are urging data sharing against information silos. But at the same time, we witness a serious monopoly of information and data, where data has become a tool for monopolizing business and demanding revenue from data providers.

There was a time when we believed that with the development of the Internet, information would be more open, its transmission would be faster and wider, and people would have richer and more convenient access to information, which would promote open and independent thinking. What we didn't expect is that the Internet information is mixed with falsehood and spam. Various AI algorithms sometimes solidify people's access to information and build a wall of thought for everyone.

AI and customer profiling are initially used for the precise marketing of goods. But when everything on the Internet is imbued with commodity and marketing attributes, the marketing of ideas and thinking becomes more mainstream. When a person accepts an idea, the algorithms will constantly send similar ideas, and sometimes more extreme and vulgar ones. Ultimately, people become more obstinate and extreme, creating unseen risks of social unrest. An economic result of the inculcation of such ideas is over-consumption and over-borrowing.

Because of the development of the Internet and the future of the digital economy, "data" has suddenly become a widely narrated term. We hear sayings like, "the one who has the data has the world". Before that, there were sayings like, "the one who has the access has the world" and "the one who has the flow has the world." Therefore, data has become an "asset."

However, there are two sides to data as an asset, just as there are two sides to technology. The good side is self-evident, as narrators have addressed extensively; the negative side is rife with the exploitation and misuse of data. Personal privacy and the right to personal data are being emphasized to the point of data sovereignty. However, I feel that the harm of generalizing and misunderstanding the concept of data may be far more than we can imagine.

A couple of years ago, some friends of mine in the smart management industry told me about the wonders of big data, bragging about how much data they already had. But I said, "As a manager, I'm fully aware of the importance of having enough information and the synergy of bridging all types of information segregation. But as a bank manager, someone who deals with international business, and especially who has worked abroad, I know better that the issues of what kinds of information you should capture and data security are more important than having enough information or breaking down information silos. If one captures the information he/she shouldn't touch and applies where it shouldn't be used, the consequences will be inconceivable."

With the adequate development of technology, mankind now realizes that people need to live in harmony with nature. Similarly, technology needs to live in harmony with humans, and technology platform companies need to live in harmony with society.

Science and technology are neutral, and it's up to humans to decide whether they are good or evil. Also, it's up to us to decide whether the result of profit-seeking capital is good or evil. It is impossible to rely solely on the self-moral restraint of the capital owners to make science and technology live in harmony with human beings and to make science and technology platform companies operate in harmony with society. We will need related arrangements of public systems and mechanisms.


What makes technology platform companies work is the effective use of big data. Therefore, to create an environment in which large technology platform companies live in harmony with society, the first step is to clarify some issues of big data.

First, the concept of "data" needs to be more clearly defined. Now "data" is a heterogeneous term, sometimes confused with "information", "material" and "intelligence".

Of course, with big data technology, there is nothing that cannot be turned into "data". However, is there any change after the "data" is collected and processed by big data technology and used as assets? What is the difference? Is the "data" an asset before it is collected and processed by big data technology, and can it be traded?

I think the differences are crucial. It is also a prerequisite for confirming the ownership of the original data, the right to use it, and future benefit sharing.

Many people believe that most data is generated by people in their transactions, life, and other behaviors; it doesn't matter who owns it, because it isn't data if companies don't collect it. Others are loudly calling for information sharing, the target of which is not these behavior data, but the original information, identity of individuals, institutions, and other kinds of related information. It cannot be denied that these are two different types of data or information. It's possible that only by discussing the two types of data separately can we distinguish the legal boundaries regarding the data.

Second, a distinction needs to be made between sharable and non-sharable data. There should be exact scope and duration of sharing for sharable data. The new legislation now in place requires a classification of data, which is entirely necessary.

At present, both "data" and "information sharing" are highly generalized. The same information or data has diverse meanings, qualities, and roles for different people and institutions. Not all data can be shared across society, hence "sharing" isn't an absolute concept. Different data need to be shared for different scopes and durations, and not all data can be made available for market transactions.

Many technology platform companies, after capturing data in the name of "sharing", have monopolized the data and refused to share with the community or even the individuals and institutions who have provided the data.

Third, data assets and data asset transactions should be approached with caution. A distinction should be made between data services and data trading. When data is referred to as "assets", it becomes broader. Because of this misunderstanding, many technological corporations seek to seize and sell all types of data by whatever means required.

There was once a market for information, consulting, and even intelligence services, but it was small in scope. Data services should have a big market space in the future digital economy and digital society. However, I believe that the question of whether data services are equivalent to data transactions should be discussed further.

Based on the clarification of data services and data transactions, it is also necessary to clarify what kind of data can be considered as assets. The aforementioned clear definition of "data", the classification, and the distinction between shareable and non-shareable data are all relevant to what data can ultimately be traded as assets.

It must be clear that sharable data isn't tradable. For example, banks send relevant credit risk data to the PBC credit system for sharing among participating banks within their business scope, but the PBC cannot sell such data as assets, and sharer banks cannot keep them and resell them for profit. Local data exchange centers that emerged not long ago have gone nowhere because what kind of data can be traded as assets have not been clearly defined.

Fourth, accounting of data assets. When data is viewed as an asset, it causes accounting problems. Nowadays, accounting generally includes overhead expenses such as information fees and consulting fees. But how can you determine a data asset's attribute? It is neither a fixed asset nor an inventory material or low-value consumables. How to measure the value of this asset? Should it be depreciated or amortized? Should the present value be calculated at a fair value? By discussing accounting methods, we're asking whether the market price of the data asset is variable. Without price volatility, market transactions are not active. How do you evaluate the data asset's quality? If data assets are like a black box or blind box as they are now, there is no way to determine their quality or value, and the market cannot function; otherwise, it will become a bizarre market similar to stone gambling. If this is the case, data assets will struggle to play a meaningful role in advancing the digital economy.

What's more, how should data assets be stored and used? Is it possible to lend or transfer it as an asset? Is the initial purchase still worthwhile if it can be lent? How does the recipient decide its worth if it can be transferred? These questions must be answered before the data is lent or transferred before it has been used or after it has been used but a backup has been made. What does unlimited data lending or transfer imply for the original data asset producer? Without resolving these difficulties, this market will never be sustainable; hence I believe that data transaction and exchange markets will be regulated before they are opened.


Having come clear about the above concepts and logic, let's look at the institutional arrangements necessary for technological platforms and the entire society to coexist in harmony.

First, view big tech monopoly rationally. The concentration of similar types of business on big tech platforms is a result of digital technology's synergy effect. With everything increasingly interconnected, the Internet of Things (IoT) is bound to serve as a platform that gathers all related industries and participants across the entire supply chain. This is exactly how it can come into full play. Most of the technological platforms today are far from mature. In the future, we will see various types of platforms emerge across different areas.

Thus, while stimulating healthy competition between platforms, policymakers must keep in mind that market concentration is not the scientific gauge to identify monopolies. Technology has two sides. Market concentration in the platform economy is not necessarily a bad thing; the real problem is whether platforms achieve such concentration with appropriate means.

Take telecommunications for example. The two sides in remote communication have to be able to have their signals connected, and so it's inevitable that people gradually concentrate on platforms that can enable the connection; and if the connections provided by different operators are incompatible, people will further concentrate on one or two platforms. If this is identified as monopoly only because of the level of concentration, and the leading platforms are forced to split to break the "monopoly", users will again gather on another platform or two as a matter of course. While the connections provided by different operators have become compatible which increased the level of competition to a certain extent, there remains the inertia among users to concentrate.

The same goes for other types of platforms including third-party payment and e-commerce platforms. So the recent introduction of new regulations prohibiting monopolistic moves such as asking businesses to make straight choices between two platforms is justified.

Second, big tech platforms should only be allowed to serve one single function. These platforms provoke worries about monopoly because they try to make use of customer resources to take over various types of businesses beyond their due scope of business, for example when an e-commerce platform engages in financial business taking advantage of their resources.

Some of the platforms charge fees for artificially rechanneling internet traffic in co-lending programs. There have been debates over whether this has raised the costs of financing for the real economy. As far as I see it, first, all internet platforms have issued loans at higher interest rates than banks over the past years. Technology has not made borrowings cheaper, but more expensive. Second, when these platforms support some of the lenders by rechanneling internet traffic, they become a new type of intermediation that bites off another portion of profits.

Platforms say they have been wronged. Is it? Not necessarily. E-commerce platforms serve as "online marketplaces", and online merchants can be deemed as their customers only when they are engaging in commodity transactions. When the merchants turn to other types of business that do not involve commodity transactions, they are no longer customers of these platforms. Think about the vendors in a real-world marketplace. When they sell stuff at their stalls, they are tenants of the market; but when they proceed with loans or payments, they become customers of banks. It's self-contradictory when internet platforms say they hope to reduce intermediations to streamline the transaction process while they constitute additional intermediation themselves.

In a digital economy, digital platforms have dual identities, as both commercial institutions and public goods. As public goods, they can only serve one single function for them to be credible. To this end, internet companies' business as platforms must be strictly separated from their other types of business. Internet companies hoping to engage in other types of business must compete on an open and fair basis under the same condition on the same platform with their peers.

Third, internet platforms must be licensed for data collection, governance, service, and transaction. These are the four main steps that data generation and application involved. Given data's special nature, we might as well examine and regulate these four steps separately.

For data collection, it's imperative to standardize the authorization process. Data, especially personal identity information, is now being collected almost anytime, anywhere by almost all institutions and APPs for unspecified purposes. Personal data collectors often claim that registration must be made on a real-name basis according to regulations. They collect users' names, ID numbers, phone numbers, facial recognition, or even bank card information.

When credit card fraud happens, banks usually think it's the customers' fault because they have failed to protect their information. But, when personal data is gathered and transacted everywhere at all times, customers just don't know how to keep their information safe. They don't even know who is having their information. Victims of such frauds are helpless. But it wouldn't be reasonable to have banks pay for the losses, either.

Thus, there must be clear rules as to the collection of what types of data must be licensed. Those without a proper license should not be allowed to collect specific data at all. Institutions shall only collect data within the proper range defined by their objectives of the operation.

Central authentication could be adopted at places where identity authentication is a must. For example, when visitors at parks swipe their ID cards or enter their phone numbers, the entrance system will automatically connect to the database of the public security system to finish the authentication. Related data collection, retention, application, or transaction shall not involve the park itself at all.

Data governance must be licensed as well. Different types of institutions should be licensed to manage different types of data.

Regarding data services, the content of data, the format of service, and the type of customers must all be clarified with strict rules and regulations. For example, credit agencies can only provide customer credit investigation services for financial or lending institutions; an industrial IoT platform can only provide data service for companies within its reach.

As for data transactions, sellers must be licensed, and we need further discussions as to whether the same applies to buyers as well. The key question here is still what types of data could be deemed tradable assets.

I have looked at these four steps separately because generally speaking, a technological platform engages in more than one of the steps, but there can be different combinations of these steps. We'd better not allow one platform to engage in all of the four steps. Of particular note, one shall not engage in data service and data transaction at the same time.

Take credit agencies for example. They can only provide data services for certain customers. They should not be allowed to sell the data they hold. Data sharing mentioned hereinbefore shall serve one single purpose of data service provision.

Some companies collect data for product development or marketing purposes. For example, carmakers collect data to improve their technology. To regulate data collection of this sort, first, it's important to clearly define the scope of data to be gathered. Carmakers could be allowed to collect driving records and road condition information, but different from navigation service providers, they don't need location data; second, data can only be used for R&D purposes, which means carmakers cannot provide data service or sell the data. Therefore, such companies can only be licensed for data collection and governance.

Fourth, put in place a scientific system for the regulation of technology platforms and the data industry. Applicable laws and rules are in the making, and researchers are having in-depth exploration. Here I only propose some general principles for reference:

To start with, with the digital economy boom, technological platforms, and the data industry will see fast development. These emerging sectors are very different from traditional ones. They are public goods and could be very risky. We may need a separate government department to place them under professional regulation.

In addition, to make the regulation professional, open and effective, it's important to have support from third-party institutions like accounting firms who could perform audits of the business, technology, and algorithms of these platforms by applicable regulatory provisions.

Besides, there should be separate rules and procedures for floor and over-the-counter trading of data assets. There must be uniform market disciplines applied by different departments to avoid disruptions and regulatory arbitrage.

Fifth, all technology platforms must be open. There should be a sound climate for technological platforms to coexist in harmony with the society at large, instead of one with monopolists and winners who take it all. There should be an institutional basis to ensure such openness and competition.

Take e-commerce platforms for example. On one hand, these platforms should serve one single function, which is to provide an efficient "marketplace" for commodity trading. On the other hand, to facilitate online transactions, the platforms should allow different payment tools, financial institutions, and other types of participants to compete on a fair basis with access to the same customer base.

If e-commerce platforms want to engage in these types of business, they need to establish a specialized subsidiary, obtain a license, totally separate the new business from its own, and compete with its peers on a fair basis. For instance, many banks have branches in shopping malls. They pay rentals to the owner of the mall, but the owner does not know which customers have come to the branches, not to mention determining the loan rate on behalf of the bank or charging intermediation fees. It's the same with ATMs at subway stations. If banks need the platforms to provide them with paid data service, then the platforms should submit data with clear customer information and the quality of the data should be identifiable. Platforms only provide data service; customers talk directly to banks about the business and the price.

Technology benefits mankind. There's no doubt that digital technology and technology platforms will benefit mankind as well. However, all major technological breakthroughs bring with them major transformations. While such transformation represents progress and advance of the human society, it does not always bring happiness and prosperity in the short run; on the contrary, it may give rise to turbulence, inequality, or even wars that could influence a generation.

How to reduce such frictions so that new technologies and emerging technology platforms could coexist in peace with society is a critical topic facing us all.