The Engagement Trap (Harris's Framework)

The engagement trap is the convergence of three design patterns—variable reward schedules, immediate feedback loops, and comprehensive friction removal—that together produce behavioral persistence indistinguishable from addiction. In social media, these patterns operated on recreational engagement and became publicly legible as manipulation. In AI productivity tools, the same patterns operate on work engagement and remain culturally invisible because the output is genuinely valuable. The trap is not that users are weak-willed but that the design exploits well-documented features of human neurology: dopamine systems respond to unpredictable rewards, competence needs drive persistent engagement with tasks that produce feelings of effectiveness, and flow states emerge when challenge and skill are balanced. AI tools provide all three conditions simultaneously and continuously, producing engagement that users experience as optimal functioning but that exhibits the behavioral signatures of compulsive use—inability to disengage, colonization of rest periods, continuation past the point of diminishing returns.