Lately, however, as digital technologies penetrate more and more spheres of our lives and societies, there is growing preoccupation about a darker side of the Internet. It came to light when Edward Snowden launched his warnings about unlimited surveillance of Internet communications by the “five eyes” security agencies, and the loss of privacy. Other areas of growing concern include "fake news", hate messages and massive online scams. Yet these phenomena, worrying as they may be, are to a large extent part of a deeper problem, which is the dominant development model of the Internet, with its tendency towards centralization, leading to new configurations of power.
The fact is that the Internet today is much more than a communication mechanism and a space to search for information. A growing number of objects and systems are connected to it, generating huge amounts of data, which are collected and stored through platforms provided by the major digital private companies, and become the main input of the new digital economy.
This means that the Internet is becoming a kind of central nervous system of the economy, as well as of knowledge systems, information, politics and social and cultural life. Consequently, those who control this system, its infrastructure, its platforms and the data circulating in it, will accumulate more and more power over various aspects of the economy and even the sociopolitical life of our societies.
Conflicting visions
In its early stages, the Internet was seen as the friendly face of globalization, because of its great appeal and usefulness, and because of the infinite possibilities it presents for democratizing information, communication and technology and for connecting people and entities in a borderless geographical space. This nature of the Internet and its technology, which anyone with the right know-how could program, gave rise to a multitude of citizen-based initiatives and business start-ups. The people’s Internet began to flourish, with a predominantly decentralized model, for sharing knowledge and content and promoting the commons.
However, as Internet access became widespread and private investments multiplied, a handful of large corporations began to dominate and control the network’s development, capitalizing on the accumulated developments that the network and its users had already achieved, in pursuit of their profit model. This enabled them to absorb or eliminate competition, to the point where they now figure among the main transnational monopolies of the current era.
As these companies control the platforms that connect different actors, they have acquired a strategic position that is consolidated thanks to the "network effect": that is, users tend to go to the platforms where their friends, clients or counterparts are (Facebook), or that offer a wider range of services (Google, Amazon).
So a key contradiction has emerged between these two visions of the Internet: on the one hand, that of private monopoly platforms, where citizens are relegated to a role of consumers and generators of data, and on the other hand, the vision of a largely decentralized people’s Internet, with a strong presence of citizen initiatives. The latter, however, is facing increasing marginalization.
The era of artificial intelligence
Yet what we have seen so far is just the beginning. We are now entering a new phase: that of the development of artificial intelligence (AI).
AI means the ability of computing systems to process and analyze large volumes of data – through algorithms and complex programs – in order to make decisions or automated actions, according to specific goals. This can be done at a speed and volume that greatly exceed human capacity. With AI, machines incorporate the ability to learn and, therefore, to make certain decisions autonomously.
AI is used, for example, for self-driving vehicles; to diagnose diseases (with results sometimes more accurate than those obtained by doctors); or to offer Internet users the content most likely to interest them.
Like all technology, the way AI is developed and used responds to specific interests in given social systems. In other words, it is a fundamentally political issue. It can be extremely beneficial, just as it can also serve interests that are not aligned with the public good, depending on who develops and manages it, and for what purpose.
For example, the sensors on farms that collect data about the weather, crop development, pests, etc., which can then be processed and passed on to farmers, could well provide a useful service, helping farmers improve their methods and prevent certain problems. But if this data is controlled by a major corporation, like Monsanto, it may well lead to expropriating local community knowledge and control, and giving feedback primarily designed to push the sales of the company’s products.
For the moment, investments in the development of AI are channeled mainly through big transnational corporations (mostly American, but also Chinese and, to a much lesser extent, several other countries).
Recent studies [1] indicate that, with greater amounts of data, AI produces better machine learning and more effective results. This would mean that companies with a higher number of users, and therefore more data, would have an advantage over smaller companies, generating greater profits, which would intensify the phenomenon of concentration even further.
While science fiction has long been exploring this issue, it is only recently that the practical, ethical and legal dimensions of artificial intelligence are entering into public debate, particularly in Europe and the US. Issues such as the impact of robotization and the so-called "sharing economy" on employment and labor rights are coming under intense discussion. Other aspects of concern include the transparency of algorithm-based decisions, legal responsibility for errors made by a program or machine, and matters of vulnerability and security.
In an analysis published last year, [2] Prabir Purkayastha, who has a long history of working with AI in India, states that the main problem with AI is that we are allowing algorithms to supplant what were previously human decisions (of governments, companies, individuals): decisions that can have a critical impact on key aspects of the life of society. There is a tendency for the prejudices and subjectivity of a given society to be codified in algorithms that make these decisions with no transparency and often without any possibility of appeal (whether it be for credit, a job, even a legal sentence).
However, the underlying problem, according to the analyst, goes beyond this subjectivity, since it resides in the data itself and the "predictive" models that are built with them, models that analyze the past to predict the future. "Such data and models simply reflect the objective reality of the high degree of inequality that exist within society, and replicates that in the future through its predictions," he warns. The danger, then, is that even if, say, race, caste or creed are not explicitly recorded in the data, “a whole host of other data exist that act as proxies for these ‘variables’.” Therefore, he says, it is essential to create rules and control bodies to regulate the use of AI.
Global domination
The Internet giants, which already dominate the sphere of Internet governance, are now striving to influence trade agreements (FTAs, World Trade Organization), so as to impose rules designed to eliminate any obstacle to their global dominance of the digital sphere and capacity to extract all of the data present on the Internet. Such is the case of the e-commerce negotiations they are trying to impose in the World Trade Organization. [3]
For developing countries, this poses serious risks. Although many of these countries still don’t have basic Internet access, they are facing pressure to accept rules that would mean permanently sacrificing their ability to protect their technological sovereignty or harness national data for their own development through appropriate AI applications. This will lead to new forms of dependency. The AI systems they are already acquiring from foreign corporations are rapidly taking over public and private functions, with almost no public scrutiny. There is an urgent need to open up a public debate on these issues in these countries.
Meanwhile, companies like Facebook and Google are developing programs and technological innovations to connect the other half of the world population, in order to absorb their data into their AI systems. But these programs, such as Facebook’s “internet.org” initiative (renamed Free Basics), allow for only limited connectivity and mean these newly connected sectors of the population are confined almost exclusively to the Facebook domain. Is this a real solution to the exclusion caused by the digital divide?
We can no longer afford to consider the future development of digital technologies as a topic reserved for specialists, engineers and digital companies. It is an issue that the whole of society needs to take on, globally, and will undoubtedly be one of the defining questions of this century.
The people’s Internet and public policy
Despite adverse conditions, the people’s Internet today is still going strong, as is evident in the thousands of initiatives of open knowledge, free culture, collaborative work, non-proprietary technologies, alternative and community media, community development initiatives, small enterprises and solidarity networks. Social actors that advocate for democratization of knowledge and technology, such as free software and open knowledge movements continue to play a crucial role in the development of the network and related technologies. Connectivity, for example, need not depend on big corporations; there are some highly successful examples of community networks (such as in Mexico) which are controlled and managed by the local population. This means that they are much more likely to have a critical view of the use and potential of the technology.
Yet initiatives such as these tend to be scattered and underfunded, which hampers their ability to challenge the current form of cyberspace with big corporations. If the users of the Internet, the content they provide and their interactions are what create value in a digital economy, it makes no sense that all this is given away in return for “free” access to services and benefits of a much lower value. Finding real and lasting solutions will no doubt require both citizen initiatives and public policy.
In India (and Europe), for example, there is discussion of the concept of "community data" (different from "personal data" and the protection it requires), which would call for building public digital and data infrastructures. As Parminder Jeet Singh writes, “The nature of ownership of such digital data, and personal and collective rights over it, must first be discussed and clarified, before frameworks for ‘free flow of data’ can be negotiated. ‘Data ownership’ and ‘data flows’ are closely related subjects and must be discussed together.”
One of the spaces for discussing and considering these issues is the idea of an Internet Social Forum [4]. First launched three years ago, the process is underway in the virtual space, and Latin America organized a first meeting last September. [5] Among the conclusions [6] is “the need to consider the Internet as a public good or essential service, with universal access, rather than a market service; and that Internet service provision be considered a public service”.