Abstract
The integration of artificial intelligence and the widespread diffusion of network architectures necessitate a revision of the categories used to interpret the public sphere and knowledge. Re-reading the Heraclitean concept of polemos, this study examines the dynamics of polarization and post-truth not as mere information anomalies, but as structural effects of digital media—characterized by immediacy, disembodiment, and the monetization of dissent. In this scenario, the uncritical adoption of generative AI risks resulting in a dangerous delegation of interpretive agency. By invoking the Hegelian distinction between what is merely "familiar" and what is truly "known", this paper demonstrates how cognitive effort and the acceptance of relational vulnerability remain essential safeguards against the performance-driven drift of technology, serving as necessary conditions for building a genuine practice of peace.
