Table of Contents
The attention economy behind AI systems
Lê Nguyen Hoàng pointed out that many AI systems today are designed primarily to maximize engagement. Recommendation algorithms on platforms such as TikTok, Facebook, or YouTube “depend on capturing human attention and monetizing it through targeted advertising. These systems are therefore extremely effective at maximizing engagement”.
He explained that these systems often amplify outrage and polarization because divisive content attracts more attention. Beyond online behavior, this economic model increasingly influences political debate and social cohesion.
From “brain rot” to writing anxiety
The lecture also explored the cognitive and psychological effects of AI systems, especially among younger generations.
The expert referred to growing concerns around brain rot, “which is the decline in cognitive abilities caused by endless scrolling and overconsumption of digital content”. Indeed, he stressed that dependence on AI-generated content may reduce concentration, critical thinking, and problem-solving abilities.
In this sense, he warned of the emergence of “writing anxiety”, where students compare themselves to generative AI tools and lose confidence in their own abilities. In addition, he highlighted the spread of AI-generated deepfakes and cyberbullying as major societal concerns.
Cybersecurity and digital sovereignty
Beyond individual impacts, Lê Nguyen Hoàng also warned about increasing cybersecurity risks linked to AI systems capable of accessing emails, files, or personal data.
A key concern raised during the lecture was the concentration of power among large technology corporations “that control much of the world’s information infrastructure, creating huge dependencies and raising questions about digital sovereignty”.
As he highlights: “It is essential to develop alternative systems that are more local, more sovereign, and more democratically controlled”. As a clear example, he referenced investigations by the United Nations and Amnesty International into Facebook’s role in amplifying hate speech during the Rohingya genocide in Myanmar.1
Reimagining democratic digital systems
Despite these risks, the lecture also highlighted possible alternatives for building healthier digital spaces.
Lê Nguyen Hoàng emphasized the importance of interoperability, meaning that content and services from different platforms should interact more openly rather than remain locked within dominant ecosystems.
He argues that: “owning a social network should not automatically grant control over recommendation systems, advertising systems, or visibility mechanisms”. According to him, interoperability could “weaken monopolies and encourage healthier alternatives”.
Furthermore, he presented the work of Tournesol, his collaborative platform based on four pillars: transparency, resistance to manipulation, privacy, and the promotion of thoughtful information. According to him, “current social media systems are optimized for impulsive behavior — endless scrolling on smartphones — rather than thoughtful democratic participation”.
The lecture further highlighted recent regulatory developments in Europe, particularly the European Union’s Digital Markets Act, which seeks to increase accountability for major online platforms and address the societal risks linked to recommendation algorithms.
Conclusion
Ultimately, Lê Nguyen Hoàng stressed that AI systems should be designed not only for efficiency or attention capture, but to strengthen democratic participation, social trust, and human dignity “by uniting societies rather than dividing them”.
read more: https://www.unesco.org/en/articles/designing-human-centered-ai-democratic-societies?hub=701
