Attention Mining

Definition

Attention mining treats human attention as a finite natural resource: extracted, refined, sold, and depleted. Notifications, autoplay, infinite scroll, and micro-interactions are methods for separating attention from its ore. The goal is not to inform but to find the next trigger that keeps the gaze on the screen.

Why it matters

The metaphor is not decorative; the operation reproduces fossil-fuel logic in the cognitive layer. Gains are collected today; damage — sleep, focus, mental health, the erosion of shared time — is amortised across time and populations that do not appear on the extractor's ledger. Attention mining is also the substrate on which algorithmic surrender runs: first attention is extracted, then the direction of that attention is set by the model.

How it appears in AI systems

In product KPIs that reward session length over task completion; in autoplay defaults that outlast the user's intention; in notification systems tuned to interrupt at cognitively vulnerable moments; in recommendation loops that harden priors rather than update them.

Examples

• Streaming services whose default is the next episode, sized to reset attention before the decision to stop.
• Social platforms whose 'unread' badge is engineered to survive dismissal.
• News feeds that surface high-arousal items in the first three positions after each launch.