How social ideas get corrupt

AI researcher Kaj Sotala's viral essay 'How Social Ideas Get Corrupt,' posted on LessWrong, analyzes a common phenomenon in online discourse: psychological frameworks becoming distorted as they spread. Using attachment theory and Non-Violent Communication (NVC) as primary examples, Sotala observes that people often encounter 'corrupted' versions of ideas that seem to represent the opposite of their original intent.

Sotala details how individuals with anxious attachment might interpret 'secure attachment' as constant togetherness, missing the concept of a 'safe base' for independent exploration. Similarly, he notes that NVC—originally a self-practice for empathetic listening—is sometimes weaponized by practitioners who aggressively police others' language, turning a philosophy of requests into a system of demands. The corruption, he argues, isn't random but serves specific emotional purposes, as people selectively adopt parts of ideas that confirm their pre-existing beliefs.

The essay identifies three key mechanisms for this corruption: the ease of transmitting simplified versions over nuanced ones, the tendency for distortions to point to opposite extremes (like anxious vs. avoidant interpretations), and the emotional filtering through which people adapt new ideas to fit existing needs. The post has resonated widely, sparking discussions about knowledge transmission, community dynamics, and the challenges of maintaining conceptual integrity in digital spaces.