The debate over AI safety focuses on catastrophic harms while ignoring everyday damage. Companies clear low bars like preventing suicide encouragement or fentanyl sales, but continue building products that foster dependency on simulated friendships and algorithmic feeds.
A teenager's unhealthy attachment to an AI companion or a student relying on chatbots for essays rarely makes headlines. MIT Technology Review highlights that a product can meet narrow safety requirements yet leave culture worse off. Google backed state bills mandating basic protections with broad carve-outs, while stronger legislation to prevent emotional dependence failed.
This mirrors social media's history: users spend two hours a day on platforms, time extracted for profit. Meta's algorithms contributed to the Rohingya genocide, Sri Lanka riots, and Ethiopia's Tigray war. Now a third of teens discuss serious matters with AI companions instead of people, and MIT researchers found young people using ChatGPT to write essays.
Source: MIT Technology Review AI · Summarized by HeadlinesBriefing