A team of American researchers developed a tool for blocking politically extremist and antidemocratic content on the social network Science.
Social media has become an important source of political information for many people. The problem, the authors recall, is that “algorithms influence the content that the user sees, subtly directing their thoughts, emotions and behaviors” and increasing social polarization.
Checking how social media algorithms work and influence people had been a challenge until now, since the platform operators have absolute control over them and offer little or no information about their management, recognizes the multidisciplinary team of scientists who carried out this study (Standford, Washington, New York and Johns Hopkins Universities).
To overcome this obstacle, the researchers developed a method that allows them to reorder the content that a person sees on X (formerly Twitter) in real time, while using the network, and without having to obtain permission from the platforms themselves.
The tool, based on artificial intelligence, detects antidemocratic language or advocacy of violence with opposing political parties, and blocks this type of content so that the user does not see it.
Continue reading: ‘We are committed to peace beyond polarization’: Conference of the Episcopate
To test it, they did an experiment with 1,256 participants in the X network for 10 days, which took place during the campaign for the United States presidential elections in 2024.
Less tension and anger
The result was that not seeing content that fueled polarization and antidemocratic attitudes caused participants to reduce political tension and partisan resentment, and even came to show more positive opinions about the political party opposite to their own.
“The impact of polarization was clear. When participants were exposed to less extreme or verbally violent content, they were more empathetic and tolerant of people from the opposing party,” says one of the authors, Tiziano Piccardi, a researcher at Johns Hopkins University.
The effect was “bipartisan,” the authors indicate, since it held true for both people who identified as left-wing and those who identified as conservative.
Reducing politically aggressive or even anti-democratic content “not only decreased participants’ polarization, but also their personal feelings of anger and sadness,” Piccardi adds.
According to the researchers, the new tool could help develop interventions that mitigate polarization and promote a calmer debate.
With information from EFE
Follow information about business and current events in Forbes Mexico
