Are algorithms a danger to democracy?
In Argentina 4.0, The Citizens Revolution (Prometeo, 2013), we propose that democracy in the digital age, increasingly influenced by new technologies, would need some kind of “mediation” of public discourse.
Algorithms are performing that task. Are they the “institutions” of digital democracy? Ultimately, new technologies deposit in them the power to structure behaviors, influence preferences, generate consumption patterns, etc.
An algorithm is the formula used by a search engine (Google, Bing, etc.) to organize the answers they will give to information requests we make. They determine the ranking of appearance for a website or news reports against a query made by the user.
The mediation of the algorithms in its current version, however, does not seem to be the type of mediation that strengthens democracy.
In 2016, when the race for the nest year´s German federal elections began, former Chancellor Angela Merkel warned of the dangers posed by algorithms, holding them responsible for distorting the perception of voters, limiting the information they received based on their previous searches in internet engines, their geographical location, etc.
In this way, she alluded to the role of the so-called “filter bubbles” and “echo chambers” of Internet search engines. “This is a development to which we must pay careful attention,” she said, because “a healthy democracy depends on people coming up against opposing ideas.”
Around those years, Cambridge Analytica´s scandal broke out, a data analysis company that contributed to the victory of Brexit by harvesting information from Facebook users (without their consent) to direct the news they received or to probe them psychologically in a disguised way. Donald Trump’s campaign (2016) was singled out for using the same tactics.
The danger posed using algorithms has grown since, along with increasingly powerful versions of artificial intelligence, based on the exponential growth of “machine learning” and the use of Big Data.
Yuval Harari defines the phenomenon as dataism or “cult of data” (Homo Deus, 2017), which turned “a limited breakthrough in the field of computer science into a world shattering cataclysm”. For dataism, “organisms are algorithms” that process data. Colonies of bacteria, forests and cities are data processing systems.
Even the ways of organizing the economy (through the market or with state planning) and political systems (democracies or autocracies) are competing data processing systems looking to predict the behavior of the systems they govern.
In a seminal book (The Age of Surveillance Capitalism, 2019), Shoshana Zuboff takes the debate further, proposing the rise of “a new economic order that reclaims human experience as free raw material for hidden commercial practices of extraction, prediction, and sale” and a “parasitic economic logic in which the production of goods and services is subordinated to a new global architecture of behavior modification”.
Just a few years after Merkel complained that algorithms chose the results that were shown to us first as a result of our Internet searches, limiting the information we receive based on our preferences, Zuboff argues that new technologies can go further. and show us information in order to modify our behaviors and determine our preferences.
Anita Gurmurthy and Deepti Bharthur warned in their “paper” (Democracy and the Algorithmic Turn, 2018) that citizens, “masterfully manipulated by data-based tactics… find themselves increasingly located on the respective sides of an ideological divide”.
More recently, professors Birgirt Stak and Daniel Stegmann (with Melanie Magin and Pascal Jürgens) published a paper that tries to answer the question stated at the beginning of this article (“Are algorithms a threat to democracy? The rise of intermediaries: a challenge to public discourse”, 2020)
They estimate there that “social networks facilitate the dissemination of false news and hate speech”, although “individuals’ exposure to these phenomena is limited”. For them, “fears that algorithmic personalization will lead to filter bubbles and echo chambers are likely to be exaggerated, but the risk of fragmentation and social polarization remains.”
It is probable that in 2023, technological progress has increased the risks signaled by these authors, who conclude their “paper” proposing a “research agenda on the governance of digital platforms”.
This is the same warning expressed in “Argentina 4.0 The Citizens Revolution”, regarding the need to find mediation mechanisms for public discourse in the digital age.
There is no doubt that digital democracy has put the very concept of citizenship at a crossroads. Autocrats have learned how to use new technologies to polarize voters and maintain power.
That´s why is key to regulate the ability of large digital platforms to influence public opinion (or at least better understand it), in order to preserve citizens freedom of choice and the healthy functioning of democracy
