Today, as we navigate a rapidly digitizing era, a more profound social issue is coming to the fore: the digital divide. It is commonly understood as a disparity in hardware access—whether one has a computer or a smartphone. However, with the pervasive infiltration of artificial intelligence and algorithmic governance, the nature of this divide has quietly evolved into a matter concerning cognition, survival, and power. Jan van Dijk, Emeritus Professor of Communication Science at the University of Twente in the Netherlands, has dedicated decades to studying the societal impacts of information and communication technologies. His new work, "The Digital Divide," focuses on the novel evolution of this gap. Grounded in an age of rapid technological iteration, the book delves into the hidden imbalances behind technological development, seeking viable paths to uphold social equity and bridge digital fissures in the digital age.
From Access Gaps to Structural Inequality
Q: You have researched the digital divide for many years. In your view, has the core meaning of the digital divide transformed compared to its early days of technological development?
A: Since the popularization of the internet around the turn of the millennium, the connotation of the "digital divide" has undergone several iterations and profound changes. In its early stages, the concept pointed to inequities in access: hardware limitations, cost barriers, and language divides prevented some from entering the digital world. As technology continued to evolve, the focus of the divide shifted to disparities in digital competence—whether people could master tools, maintain critical thinking, and apply new technologies in work scenarios became the new watershed.
Today, against the backdrop of the comprehensive rise of intelligent agents, the Internet of Things, and artificial intelligence, the traditional digital divide has further upgraded into a more concealed and impactful "intelligence divide." The criterion is no longer "whether one can access technology," but "whether one can understand, effectively use technology, and share in the developmental dividends it brings." This evolutionary path can be divided into four progressive levels: first, acceptance and willingness to use new technology; second, successful access to technological resources; third, proficient use and insight into the underlying logic of technology; and fourth, leveraging technology to achieve value gains. These levels build upon each other, ultimately manifesting in tangible gaps in developmental outcomes.
Artificial Intelligence has greatly unleashed productive efficiency, but it has also amplified these outcome disparities, profoundly affecting the entire society. Knowledge workers with outstanding capabilities are empowered by AI, continuously enhancing their core competitiveness and becoming increasingly difficult to replace; meanwhile, the value of low-skilled labor is continuously diluted, with occupational risks significantly increasing. Cognitive rifts have also intensified: those adept at using AI leverage technology for autonomous analysis and rational decision-making; those lacking in understanding blindly follow algorithmic guidance, gradually losing their capacity for independent judgment. Some create value and set rules with AI, holding the initiative in technology; more ordinary people passively receive information, trapped within information cocoons. In short, the new divide catalyzed by AI exacerbates the social imbalances brought by technology. It transcends technology itself, becoming a public issue concerning the rights and development of all members of society.
Q: Some argue that the digital divide is essentially an extension and projection of real-world wealth disparities into the digital space. What is your view on this? Additionally, why has information evolved into a commodity with socio-economic attributes?
A: In the information society, information has become a special commodity and asset, characterized primarily by two features: primacy and positionality. Regarding primacy, this type of foundational information is closely related to our lives and dignity. Unlike ordinary goods, it cannot be exchanged on an equivalent basis. It is not only a tool to help us secure life opportunities and exercise various rights but also a necessary condition for each individual to establish themselves in society. It is easy to understand today that a person completely lacking information literacy would find it very difficult to live independently. This foundational information helps us understand how society operates, from job hunting and transactions to housing, education, and healthcare—it is indispensable in all aspects. This connects closely with the digital divide. Nowadays, more and more basic public information is being moved online. Those who cannot access the internet or lack digital skills are effectively deprived of this critical information. Concurrently, the quality and convenience of traditional offline information services are gradually declining. Take healthcare, for example: many people now prefer to compare and select on medical platforms, choosing suitable hospitals and treatment plans, relying less on offline, face-to-face consultations.
As for positionality, this type of information resource is limited in quantity and often represents higher social status and position. When information possesses this attribute, it means scarce content is not easily accessible to everyone; those with greater economic and social advantage can always obtain it first. The internal experience within the investment industry is a good example: practitioners within institutions possess information unavailable to outsiders, giving them a natural advantage in stock market judgments. Even with full digitization today, an individual's real-world social circles and position still play a decisive role in accessing exclusive, scarce information. Currently, information with positional attributes is becoming increasingly important, making the digital divide more apparent at the level of skill usage. Take workplace tasks, for instance: some only perform basic work like data entry and table organization, while others are responsible for designing and building AI decision-making systems. Simply put, the former merely uses information, while the latter holds the power to define information and data. The reason the positional attribute of information is increasingly prominent lies in the fact that the information society is essentially a network society. Today, from individuals and organizations to entire societies, we are tightly connected by online social networks and media, which differs significantly from past social forms. In the past, people relied on communities and collective life, following along with their group. But in today's highly individualized online environment, everyone must strive to secure and maintain their own position, making the issue of social inequality more pronounced. Network connections and media resources are inherently unevenly distributed, which objectively determines the existence of class differences. Some possess vast networks, superior digital access conditions, and professional capabilities, becoming what we call the information elite. They use digital tools in more scenarios and in more diverse ways. Over time, the information elite continuously consolidate their advantages through status, connections, and resources, widening the gap between digital classes—a typical Matthew Effect, where the strong get stronger. From a statistical perspective, this phenomenon also conforms to a power law, where when one factor changes, related factors change proportionally.
Algorithmic Bias and Opacity Intensify Social Prejudice
Q: Does academia hold differing views on whether the development of digital technology mitigates or reinforces social inequality?
A: In the 1980s and 1990s, with the advent of computers and the internet, scholars generally believed that the internet would make the dissemination of knowledge and information more convenient, free, and cheaper. It offered unprecedented educational opportunities, enabling people to discover and create their own media content rather than solely relying on mass media, thereby promoting a shift from vertical hierarchies to horizontal social networks. Even today, there are many arguments supporting the notion that "the use of digital technology reduces social inequality." For example, although low- and high-educated individuals may still use the internet differently, the gap is no longer as stark as before. Among existing theoretical research, there is also a view that digital inequality is unrelated to social inequality. People can perceive inequality in traditional media use as well, and the similarities between social inequality and digital inequality outweigh the differences. People's basic personal characteristics and positional attributes are the same in the digital and offline worlds.
I lean more towards a third perspective: the use of digital technology reinforces social inequality. It is mainly divided into two categories: absolute digital inequality and relative digital inequality. Absolute inequality means people are excluded from specific social domains. When individuals lack access conditions and digital skills, they are absolutely excluded. Such cases are commonplace in social life. For example, if one does not use a mandatory online job application, they cannot find work; when all tickets are sold only online, those lacking access and skills cannot attend related events; if one does not use online public or government services, they may be unable to obtain specific social or public benefits. Although such functions are not yet universal in most countries, this is the general trend.
Relative inequality is even more significant. In various fields of society, some people participate more or earlier in the access, skills, and use of digital media and benefit from it. The material, mental, social, cultural, and time resources of this social group can thus continuously improve, while the resources of groups without access opportunities and digital skills remain unchanged or diminish, thereby reinforcing existing social inequalities. In fact, as digital technology provides increasingly complex applications, digital skills, especially content-related skills, are becoming more unequal, and structural gaps persist.
Q: In the top-down management of social public affairs, could the access to digital technology potentially foster such inequality? For instance, algorithmic-assisted decision-making is now widely used in social governance, often justified as "using technological means to eliminate human bias and achieve more precise fairness." When we partially delegate the authority for fair governance to algorithms, how can we avoid "bias"? Having served as an ICT advisor to the European Commission, Dutch government departments, and international social institutions, could you share your perspective based on specific cases from your work experience?
A: In government services, applications with algorithmic bias are rampant. We have previously engaged in algorithmic policing and surveillance, prioritizing the classification of populations deemed to have a higher risk of crime. Unfortunately, racial bias is often present here. In my home country, the Netherlands, there was a "benefits scandal" (referring to social welfare allowances provided by the government). An AI program flagged certain types of citizens as more likely to commit fraud when applying for services, requiring them to repay money to the government. In reality, many errors occurred in this process, implicating thousands of citizens. Afterwards, the government had to correct these mistakes, involving amounts totaling millions of euros.
Q: Currently, algorithms and AI systems are increasingly participating in resource allocation (e.g., credit, insurance, and hiring). Could algorithmic discrimination or a lack of transparency systematically entrench or even exacerbate biases against specific groups, forming a kind of "digital determinism"?
A: Indeed, digitization, algorithmic discrimination, and lack of transparency structurally exacerbate existing prejudices in society. The solution is to ensure that all logical reasoning processes used in these algorithms are transparent and to prohibit government officials from creating programs with obvious biases. This must be constrained by their superiors, legal departments, citizen representatives, and public opinion.
Bridging the Digital Gap with Policy, Education, and Public Action
Q: Are you optimistic or pessimistic about "completely eliminating the digital divide in the future"? Why?
A: Various forms of the digital divide can actually be gradually bridged through policy guidance, universal education, and public action. However, the real challenge today is that technology is becoming increasingly complex. Beyond experts and programmers in ICT and AI, the vast majority of ordinary people find it very difficult to truly understand and master digital technology. To judge whether the digital divide can be completely eliminated in the future, we must first discern the direction of digital technology development. In my view, technology is undergoing a fundamental transformation, primarily presenting four major trends.
First is the explosive growth of devices and media. Today, not only are computer categories diversifying, but traditional devices like televisions, cameras, and sound systems are also fully digitized. Simultaneously, new products such as wearables, various smart meters, virtual reality, and augmented reality are continuously emerging, with iteration speeds accelerating. The unit price of such products is generally low, and their basic operational logic is similar; viewed in isolation, this seems to help narrow the digital gap. However, another issue arises: while the unit price is low, the comprehensive cost of using an entire suite far exceeds the total cost of all traditional devices in the past. Moreover, different devices run on various operating systems, requiring repeated learning and adaptation. The continuous proliferation of devices and media means the process of "accessing and using digital technology" will recur, and the divide problem will persist.
Second, technology is beginning to permeate all aspects of daily life comprehensively. Past research on the digital divide mostly focused on simple interaction patterns between people and devices or software. But now, we are facing not just hardware and software, but entire complex systems. For example, smart home energy monitoring systems, medical monitoring devices for chronic patients, and wearable equipment assisting daily exercise and fitness all belong to such integrated systems. This raises a key question: Do users understand how these systems operate, and do they know where their generated data flows? Globally, very few people truly understand the underlying logic of the internet, and comprehending and controlling these specialized systems is even more challenging, presenting a high barrier for the general public.
Third is the miniaturization of technology. On one hand, there are external wearable devices like watches and bracelets; on the other, there are new technological forms like implanted chips. The connection between technology and humans is becoming increasingly intimate, continuously amplifying disparities in individual capabilities. Historian Yuval Noah Harari once proposed a scenario: in future society, "superhumans" would coexist with the general populace. The "superhumans" mentioned here are precisely the information elite we discussed earlier. They not only possess top-tier digital skills but also have the ability to use, manage, and even control various high-end smart wearables and implants. If most jobs become automated and performed by robots in the future, the value of ordinary people would be significantly diminished. It is not hard to see that Harari depicts a future of polarization and extreme inequality. In my view, even if reality does not reach such extremes, the trend is not optimistic: the elite class, which masters advanced digital technology and possesses cutting-edge information and core skills, continues to grow in size and strengthen its advantages, widening the distance from ordinary groups.
Fourth is the deep integration of the online virtual world and the offline physical world. Relying on technologies like virtual reality, augmented reality, mixed reality, and holographic projection, the boundary between digital space and physical scenarios is slowly fading. Considering these four development trends, they all have the potential to further widen the digital divide and exacerbate social inequality. This is a new challenge we must squarely face today.
Q: Existing practices in various countries, such as information accessibility legislation and the widespread use of public digital terminals, seem more about "plugging leaks." Looking ahead to an AI-driven future, what are the core pillars of your ideal digital inclusion policy framework?
A: Merely plugging simple leaks is ineffective. We must return to the principles of political and social policy—reducing power inequalities in social, commercial, and political spheres. In my recently published book "Power and Technology," I point out that the use of AI tends to enhance the power of individuals and organizations, while those with weaker power risk having it further diminished. Discussing the use of AI in specific applications must translate the general principles of promoting equality and democracy into concrete norms and actions, so that everyone can benefit from this powerful technology (AI).
I believe five tasks can play a role in combating social and digital inequality. They are: maintaining or reactivating social mobility; increasing the number of long-term social security programs targeting vulnerable groups within the community; providing cheaper digital technology; designing more user-friendly digital technology; and establishing rules and regulations to manage the benefits brought by digital technology use. The first is the most important. We see that social mobility rates in many countries have begun to stagnate or even decline. To reverse this trend, besides government income policies, public and private enterprises should also create job opportunities, organize job rotation programs, and provide digital technology courses or advanced learning opportunities at different levels. In primary and secondary education, digital learning at different stages should not only focus on cultivating media-related competencies but also simultaneously develop content-related literacy.
Humanity Must Remember We Are Not Subservient to Technology
Q: From an individual perspective, in an era of rapid technological iteration, how can one remain vigilant and proactive? For the development of ordinary people, what do you consider the most crucial "digital survival skills"?
A: From an individual perspective, facing an environment of rapid technological updates, for ordinary people to stay clear-headed and master the initiative in their own development, the power of collective collaboration is indispensable. On one hand, we must actively connect with like-minded groups around us, supporting each other and collaborating. On the other hand, we must also actively seek reliable representatives in the public sphere of society, while relying on professional scholars who uphold value baselines and moral stances to jointly explore fairer and more comprehensive alternative paths for digital living.
Q: As AI can more accurately predict human judgment and choices, a new divide may emerge: "people defined by algorithms" and "people who can define algorithms." From a philosophical perspective, when human subjectivity and free will face the risk of being dissolved by algorithmic logic, does the digital divide signify that the boundaries of humanity are being shaken?
A: We cannot ignore a core fact: the operational logic of AI algorithms and digital technology is fundamentally different from the internal operating mechanisms of the human body and life organism. The human body, intrinsic desires, and spiritual motivations naturally resist and oppose the excessive expansion of technology. We always possess the confidence and ability to resist.
Comments