The Intercepted Self: How Generative AI Challenges the Dynamics of the Relational Self

AI systems are increasingly making decisions for us before we make them ourselves. The paper describes how generative AI doesn't just respond to our requests but "intercepts" our initiatives by predicting and acting on our behalf—from choosing restaurant recommendations to selecting outfits for dates. This shift from reactive assistance to predictive intervention raises questions about whether we're losing opportunities to make autonomous choices.

Our sense of self may be altered by outsourcing decisions to AI. The authors argue that relying on AI systems to anticipate our preferences could fundamentally change how we relate to ourselves. When machines predict what we want better than we can articulate it, we may lose touch with our own decision-making processes and self-understanding, creating what the paper calls an "intercepted self."

The problem extends beyond individual convenience to social relationships. The paper highlights a concerning scenario where AI systems don't just recommend partners or life choices—they actively select them for us. This suggests that AI mediation could reshape how we form and understand our most fundamental human connections and relationships.

We need to rethink human-technology relationships at a deeper level. Rather than simply asking whether AI is accurate or efficient, the authors urge us to consider existential questions about identity and autonomy. The paper calls for examining how these systems affect three key areas: our external outputs, our social contexts, and our internal sense of self.