People use their smart phones on the Subway Line 8 in Beijing, Sept. 21. Photo: Chen Mirong/CSST

In recent years, the concept of the digital society has become the fundamental backdrop for researchers to conduct their studies, and it is also profoundly transforming the interpersonal dynamics between individuals. The focus of research has shifted from the question of whether to use digital technologies to determining which technologies are most suitable for specific scenarios. People now heavily rely on e-commerce platforms for their everyday economic activities, and their interactions with the government often take place through government apps. As a result, a majority of interpersonal relationships are established and maintained through messaging apps and social media platforms. The increasing volume of social interactions facilitated by technology has resulted in a social dynamic that can by encoded by computer systems.

Encoding social relationships

The encoding of social relationships involves digitally recreating or representing particular elements of social interactions. This process transforms the contextual dynamics of social relationships into a phenomenon of technological-social integration. While these platforms still embody the fundamental principles of human social dynamics, such as establishing connections by clicking “follow” or cultivating a wide network of friends akin to being a “social butterfly,” they also encode the fundamental logic behind these actions. For instance, in order to view another individual’s activities, one must click “follow,” and engaging in real-time communication necessitates the recipient’s confirmation of a friend request. Additionally, adhering to the social platform’s norms is crucial for gaining platform-generated traffic.

To understand the problematic aspects of “encoding” social relationships, it is important to consider two key features.

First, technology’s regulation and restructuring of social relationships has become the norm. The 51st Statistical Report on Internet Development in China, published by the China Internet Network Information Center, reveals that Chinese internet users dedicate an average of 26.7 hours per week to online activities, equivalent to nearly 4 hours per day. When we consider the standard 8 hours for work and 8 hours of sleep, it becomes evident that a substantial portion of people’s free time is spent online. This highlights the profound integration of technology and the digital extension of social relationships. Therefore, discussions about the future forms of social relationships should gradually shift from debating the presence or absence of digital technology to examining the specific types of digital technology involved. This will allow for a deeper analysis of the methods and directions digital technology takes in encoding social relationships. 

Second, social actors’ recognition of encoded social relationships is a prerequisite for analysis. Whether a social relationship has been encoded is not based on whether it can be transformed into code on a digital platform, but rather by whether the encoded relationship truly functions among multiple actors. For example, some districts attempt to implement digital governance over rural areas by creating a “digital twin” of daily practices through digital technology. However, if villagers have no intention of using this technology in their daily lives, the digital governance system is useless. 

The varying degree to which individuals accept the encoding of social relationships can also lead to discrepancies in research interpretations regarding the same types of interactive social relationships. Among “digital natives,” instant messaging apps and social media platforms are considered integral to genuine social interactions, leading to the evolution of complex online social etiquette. However, those who grew up before the digital era generally agree that online social activities are merely a form of virtual relationship. They tend to prefer maintaining social relationships through face-to-face interactions, emphasizing offline social etiquette over online social norms. Therefore, all encoded social relationships must be discussed in the specific context in which they occur.

Three mechanisms 

One of the consequences of the encoding of social relationships is the transformation of social relationships into action patterns that can be comprehended by digital technology. This raises the question of how technology expands its sphere of influence over relationships. This expansion occurs through three mechanisms. 

First, the structure that digitization enables easier comparison between different types of social relationships. Compared to offline social interactions, which don’t leave clear traces, relationships established on digital platforms exhibit more evident and standardized characteristics. For example, we can compare the influence of public figures by measuring their number of followers, click rates, retweets, and so on. By reviewing the number of interactions and their frequency on instant messaging platforms, the length of chats, whether a contact is “pinned” to the top of one’s chat list, and other indicators, we can measure interpersonal distance. Although each of these indicators may have certain issues of reliability and validity, the structured data formed through encoding consistently provides various new indicators, which mutually support each other to a certain extent. This structural trend not only measures social dynamics, but also marks a social and cultural consensus, which can be used as an important indicator to measure social relationships, and predict how follow-up social interaction activities may unfold. 

Second, digitization allows social interactions to be transmitted quickly. On digital platforms, all social interactions and relationships can be transformed into data streams that can be transmitted, stored, analyzed, processed, and matched in an instant. Digital social relationships are no longer confined to a fixed geographical location or time, thus making a wider range of interpersonal relationships possible. Niche social interests, which were originally confined to the narrow matching space of local face-to-face interaction, have now formed large-scale online subculture groups through the form of data. In recent years, as young people have grown more diverse, dating software and match-making websites have gradually become one of the main ways for young people to develop romantic relationships. This also indicates that multiple value orientations require the support of algorithms to encode social relationships. 

Finally, encoding embeds social relationships in a wider range of scenarios. When social relationships are abstracted into specific codes, it means that the relationships can be placed in broader social contexts that are also encoded. Using the promotion of public service activities as an example, when social relationships primarily unfold on within instant messaging apps, the platform itself can encourage users to forward information to others in their online social network. The platform can also create interesting online public service activities to keep users engaged in charitable efforts. An important prerequisite for the success of these actions is that relationships must be transferred to the digital world, and once they are in the digital world, it becomes easy to move social relationships into other scenarios. 

Virtual ‘social relationships’ 

With the development of digital technology, a growing number of relationship forms can be encoded, leading to another social consequence — the gradual detachment of physical people from social processes. In certain contexts, individuals do not need to interact with real people to experience a “social connection.” With the advancement of artificial intelligence technology, the degree of authenticity in virtual “social relationships” has further improved, accelerating the replacement of physical people with AI chatbots. This can be examined along the following three aspects. 

First, virtual relationships demonstrate higher levels of social affinity than average people. The degree of socialization between individuals is different, and people can only achieve limited socialization through physical processes. With the development of AI technology, the degree of socialization, or social affinity, offered by chat-bots that are trained on real interactive data online may gradually exceed the social reach of real humans. When people become aware of the social risks of AI, such as discrimination, adjustments are made to the programming, which will further improve the sociability of AI digital tools. Therefore, in the near future, being in a virtual relationship may be more comfortable than interacting with a real person. 

Second, virtual “social relationships” have greater capabilities for “socializing.” Compared to interpersonal interactions where interactants may forget important information, virtual “social relationships” have the ability to record more micro-level information during social interactions, such as consumer preferences, user habits, and so forth. By analyzing users’ trace data, optimization and adjustments can be made to the direction of social relationship interactions. Eliza, a chatbot from the 1960s, was already able to select appropriate conversation content based on prompts set by its programmers, so the real people it spoke with felt it was an empathetic listener. In June 2022, Google’s LaMDA conversation program also had the ability to extract information from a chat partner’s language patterns to form an interactive “conversation,” convincing engineers that the program had a soul. In terms of processing interactants’ information, virtual “social relations” may comprehensively overtake real interpersonal interactions. 

Third, virtual “social relationships” are more likely to generate technological risks that are beyond the imagination of current social perceptions. Based on the above capabilities, virtual “social relations” can provide 24-hour uninterrupted interaction, highly personalized experiences, and rich social scenes. They can also carry out a series of actions centered on the interactant. This flexibility, variety, and power accelerates the process of people shifting their emotional dependence from physical to people to virtual entities.  However, how this virtual “social relationship” will profoundly impact the social forms that have revolved around physical individuals for thousands of years, and what social risks and structural changes it entails, is a question that the current social science knowledge system has yet to provide an accurate answer to.

It is true that digital technology is still not capable of recreating digital people with human warmth and logic. However, the continuous 30-year process of encoding social relationships has enabled digital technology to empower a considerable portion of the population to consciously or unconsciously perceive interactions with virtual beings as real social relationships. Humans and machines have moved toward a state of “mutual interdependence.” Questions of how to properly deal with non-real social relationships and predict the social risks and structural changes derived from these social interactions will be important topics for future research. 

Zhang Shuqin is an associate professor at the School of Sociology and Psychology at Central University of Finance and Economics; Song Qingyu is a post-doctoral candidate at the School of Public Management at Hohai University.

Edited by YANG XUE