Algorithmic Incentive Misalignment

mechanism

On a free, ad-funded platform, the algorithm’s working objective is not your flourishing but your continued attention: more time creates more opportunities to show you ads.

The feed presented as a personalized service is working for someone else. Its success is measured partly by how much longer it can keep you there—and how many more ads fit into that time.

E1

Follow the paying customer

When access is free and advertising funds the platform, your attention becomes an input to revenue. The system is therefore rewarded for extending sessions and encouraging return visits. This is Engagement Optimization: the algorithm need not dislike you or understand what improves your life; it only needs to learn which selections keep you consuming. The misalignment sits between the outcome you may want—useful information—and the outcome the platform can monetize—more time available for ads.

E1

An incentive is not a verdict on every recommendation

The business model explains the system’s directional pressure, not the value of each item in your feed. Some recommendations may genuinely help you. The sharper claim is that helpfulness is not the objective demonstrated by the supplied evidence; continued attention is.

E1

Price the feed in minutes and ads

Tomorrow, before opening a free platform, choose one task and a fixed exit time. When the timer ends, count whether you completed the task or merely remained available for more advertising; then close the app regardless of what the feed offers next.

E1

Episodes that teach this