Internet social mentality mapped in AI era

By MA XIAOYUE / 07-04-2024 / Chinese Social Sciences Today

FILE PHOTO: Technological innovations have strengthened control over information and interactions, changing public behavior online, ultimately transforming the internet social mentality.

The “internet social mentality” refers to the aggregate of culture, emotions, values, and behavioral tendencies present in the network society during a given period. This social mentality is often characterized by a strong sense of right and wrong and intense emotions, which manifests in digital personas which might be active and hard-working, rational and calm, or low-key and indifferent. Other popular forms of social expression include collective resentment, impulsive utilitarianism, and anxious pessimism.

Today, under the influence of digital intelligence technologies from big data to blockchain, and now the addition of large language models, the cyber world has achieved an intelligent transformation. Yet new technologies have introduced new uncertainties, making the network environment in the digital age unpredictable and complex. Individuals find themselves gambling in an expansive new game of opportunities and risks. Under this influence, the public’s social mentality has changed online.

In this new digital landscape, sociologists are mapping the communication patterns of the internet social mentality in China, to foster a positive, harmonious online space with a clear and invigorating atmosphere. This goal has become a crucial component of improving China’s social governance system.

Real landscape and root causes

The public’s internet social mentality stems not only from public reactions to life experiences and social issues but also from changes in the online information environment. The latest social mentality survey report shows that a hard-working spirit remains the main theme of China’s current online world, but negative mentalities such as passivity, anxiety, and pessimism have captured a significant proportion of the population. In the virtual world, people often express themselves more freely, unrestrained by real-world norms. Therefore, different characteristics are displayed over time, reflecting changes in the online environment.

Technology is a major driver of social change. Advances in big data and Large Language Models (LLMs) have drastically transformed the online environment. The deep datafication of society has altered the underlying logic of human survival and development. Individuals are becoming increasingly transparent online, while virtual interactions occur with intelligent machines deconstructing traditional social structures and causing a certain bias within the public’s internetsocial mentality. 

In the early stages of internet development, social media platforms like Weibo, WeChat, Tieba, and Zhihu disrupted traditional media’s information dissemination methods, achieving a degree of decentralization. Social networks, by reconstructing the information environment, reshaped public interactions and information behavior, which had a profound impact on people’s cognitive structures and value systems. During this period, the court of online public opinion was complex, however, despite the chaos, it had a positive effect on public emotional expression and social supervision.

With the widespread application of digital intelligence technologies, technology platforms have broken the previously diverse and scattered landscape of public opinion. By technologically reshaping the central authority of information dissemination, they have strengthened vertical transmission from media outlets to users. LLMs, as a cornerstone of the digital intelligence era, have led to increasingly sophisticated data analysis and recommendation mechanisms. By aggregating behavior and analyzing user-item or user-information interactions, individual interests and preferences are fully deconstructed, exposing individuals to the plain sight of technological companies and platforms. Based on this, intelligent applications can precisely package information according to users’ profiles while shielding unfavorable information.

The digital intelligence era represents an “interaction revolution” catalyzed by technology, ultimately reducing and alienating people-to-people exchanges. This alienation is not caused by common issues such as digital anonymity but by the technical characteristics of the internet. In this setting people neglect traditional social behavior norms, exacerbating indifference in real-world interactions. Technological innovations have also strengthened control over information and interactions, which changes public behavior online and thus transforms the internet social mentality.

Risks to society

As the core driving force of the new wave of technological revolution and industrial transformation, digital intelligence technology profoundly affects the international landscape and accelerates humanity’s entry into the intelligent era. While new technologies greatly facilitate human work and life, enhancing efficiency and accuracy, they also represent a “double-edged sword” for comprehensive human development.

As rivalries intensify among major world powers, generative AI might embed ideologically relevant content during key points in history or major events, creating a “pseudo-opinion climate” aimed at influencing public political judgments. In addition, AI might push biased content to targeted individuals in the long term. Through big data analysis, AI can easily determine an individual’s political bias and use this information to set agendas. When it is necessary to guide the public to ignore certain topics or distort perceptions, AI can flood the web with redundant information, interfering with the spread of mainstream ideology. Perceptions, cultural concepts, and social values represented by such data become entrenched through public opinion collisions, leading to a narrowing of public understandings and a fragmentation of values, forming a relatively extreme internet social mentality.

On a societal level, if intelligent production resets social structures, it induces group anxiety. We see this when the leap in productivity brought about by technological advancements adversely affects employment. Rapid technological development, while continuously raising people’s living and consumption expectations, also has significantly harmed various industries. 

A 2023 report by Goldman Sachs indicates that an estimated 300 million jobs globally will be replaced by generative AI, with about two-thirds of jobs in Europe and the United States being affected to some degree. As a production tool, AI’s massive computing power integrates vast human social resources, resetting economic benefits and social resource allocation, while also strengthening control over information flow and increasing the likelihood of social stratification. If not restrained, technological capital can easily siphon wealth from the people, leading to severe backlashes in public sentiment and exacerbating social conflicts. Although new technologies have mobilized a competitive spirit, many individuals face anxiety over falling behind if they fail in this technological competition, creating an atmosphere of widespread anxiety.

On an individual level, obstacles to horizontal interactions lead to changes in the online social mentality. Over-reliance on the creative functions of generative AI will continually enhance AI’s subjectivity, transforming technology from a tool for innovation to innovation itself. If this trend continues, human creativity and problem-solving abilities are likely to diminish, ultimately becoming subservient to technology, with the overall social mentality trending toward hedonism. 

Furthermore, as human-machine anthropomorphic interactions become deeply embedded in society, the normal amount of time spent on personal interactions is imperceptibly compressed. Studies show that frequent horizontal communication leads to the dissolution of hierarchy, where public gains in new perspectives and ideas erode acceptance of authority. Excessive human-machine interaction, and the sleek packaging of entertainment information, reduces the time spent on horizontal communication, while vertical dissemination from mass media becomes the main channel for obtaining information, indirectly strengthening platform control over information. 

At the same time, due to AI’s enhanced functions in virtual dialogue, information gatekeeping, and public opinion manipulation, the public cannot express overt dissatisfaction online. Thus, angry and resentful emotions are being replaced by indifference or sarcasm and widespread negative emotions are being suppressed, left to simmer below the surface. Over time, the public may either indulge in quiet sarcasm or resist information control through negative online practices like “letting it rot” or “lying flat” to the point that despondent netizens won’t be bothered to make an effort.

Communication rules

The internet social mentality is the primary manifestation of social consciousness in cyberspace, and is characterized by the most primitive and pure psychological experiences and emotional expressions, often featuring anti-authority, anti-intellectual, and rebellious rhetoric. In the past, mainstream platforms have launched a series of paternalistic lectures to reform negative mentalities, an approach generally scorned by the general public. Digital intelligence technology now offers new pathways. 

First, social bots can be utilized to establish multi-level communication networks. Research from the University of Southern California shows that 9%–15% of active users on Twitter are bots rather than humans, and this number continues to grow. In reality, social media bots are active across major platforms domestically and internationally as widespread information dissemination tools. Due to advancements in AI technology, social bots appear more human, making it difficult for people to know whether they are interacting with a machine or another human. In response, resource coordination thinking should be strengthened to harness this uncontrolled reality and use technology strategically to benefit the public. With generative AI, we can come up with more attractive, diverse, and valuable cultural products that resonate with broader audiences, to foster a positive ethos.

Furthermore, an intelligent recommendation mechanism can be used to create a positive online information environment. Changes to the internet social mentality exhibit significant community characteristics, with some significant variations in different social circles and information environments. This process intertwines emotion and rationality, due to the close integration of individuals and society. Intelligent recommendation systems can act as “safety valves” to alleviate group anxiety to some extent, effectively fostering a positive social mentality through emotional regulation. The application of digital intelligence technology can assist in screening negative content and push corresponding positive content to different groups, building a network of positive interactions and enhancing the public’s online literacy.

Finally, an intelligent monitoring system should be established. The representation and evolution of the internet social mentality serves as a barometer for public perception of environmental changes. Therefore, to effectively build a harmonious online environment, the prediction of network risks should focus on negative changes in the public ethos, where intelligent monitoring and AI emotion recognition should be in place. By fully leveraging deep learning technology to uncover hidden attitudes and emotions online, targeted decision-making and strategic intervention can be conducted, enhancing the effectiveness of civil communication and ensuring social prosperity and stability.

Ma Xiaoyue is a professor from the School of Journalism and New Media at Xi’an Jiaotong University.

Edited by YANG XUE