Přednáška: Undermining Democratic Deliberation? How Engagement Algorithms Hijack Epistemic Agency
Zveme Vás na přednášku s názvem Undermining Democratic Deliberation? How Engagement Algorithms Hijack Epistemic Agency, kterou pronese Dr. John Dorsch v pondělí 8. prosince 2025 od 13:00 na FHS UK v místnosti 2.31.
Does artificial intelligence (AI) threaten deliberative democracy? This paper argues that it does, not because it isolates us in “filter bubbles,” but because it hijacks the very capacities through which we evaluate what to believe and how to judge others. While early accounts blamed algorithmically induced isolation, we show that the real danger lies elsewhere. By optimizing for engagement, AI-driven recommender systems not only make social media psychologically addictive, they also optimize for cognitive fluency, which in turn produces spurious metacognitive feelings of confidence. These feelings lead users to mistake fluency for justification, generating unwarranted certainty both in their beliefs. Drawing on research in metacognition and epistemology, we argue that this process constitutes a form of epistemic hijacking: AI reconfigures the mechanism by which judgements are evaluated, producing a distorted sense of epistemic reliability and intellectual arrogance. We conclude by outlining how AI could be redesigned to counteract these effects and repair the evaluative conditions on which democratic discourse depends.
Zveme Vás na přednášku s názvem Undermining Democratic Deliberation? How Engagement Algorithms Hijack Epistemic Agency, kterou pronese Dr. John Dorsch v pondělí 8. prosince 2025 od 13:00 na FHS UK v místnosti 2.31.
Does artificial intelligence (AI) threaten deliberative democracy? This paper argues that it does, not because it isolates us in “filter bubbles,” but because it hijacks the very capacities through which we evaluate what to believe and how to judge others. While early accounts blamed algorithmically induced isolation, we show that the real danger lies elsewhere. By optimizing for engagement, AI-driven recommender systems not only make social media psychologically addictive, they also optimize for cognitive fluency, which in turn produces spurious metacognitive feelings of confidence. These feelings lead users to mistake fluency for justification, generating unwarranted certainty both in their beliefs. Drawing on research in metacognition and epistemology, we argue that this process constitutes a form of epistemic hijacking: AI reconfigures the mechanism by which judgements are evaluated, producing a distorted sense of epistemic reliability and intellectual arrogance. We conclude by outlining how AI could be redesigned to counteract these effects and repair the evaluative conditions on which democratic discourse depends.
| Začátek akce | 8. prosince 2025 |
| Konec akce | 8. prosince 2025 |
| Druh akce | Přednáška |
| Místo konání akce | FHS UK, místnost 2.31 |
| Cílová skupina | Akademická obec i veřejnost |
Fakulta humanitních studií UK
