Most people who have lost an evening to a feed know the sensation: the thumb moves on its own, and the hour dissolves into a scroll that felt pleasant at each step and miserable in aggregate. That gap between moment and total is exactly where two old ethical systems part ways over the attention economy. 9 min read.
A Kantian diagnoses engagement design as a failure of respect, because it treats users as instruments for someone else's revenue and routes around their rational consent. A Benthamite diagnoses it as a failure of arithmetic, because minutes spent on an app can count as pleasure in the ledger while the pains of compulsion, regret, and displaced life are never entered.
Two Ways of Keeping Accounts
Kant's test for a maxim is whether it could be willed as a universal law without contradiction, and whether it treats rational agents as ends in themselves rather than merely as means. Bentham's test is quantitative. Actions are right in proportion to the balance of pleasure over pain they produce, summed across everyone affected, and the sum is to be taken seriously in its parts: intensity, duration, certainty, propinquity, fecundity, purity, and extent. The two frameworks start from different questions. Kant asks what kind of act this is. Bentham asks how much good or harm it yields. The attention economy, a system whose product is the user's unbroken gaze, tests both questions at once, and the results are not the same.
The Feed as a Problem of Standing
Take a variable-reward notification system, the slot-machine mechanic that Nir Eyal popularized in Hooked as a model for habit formation. A Kantian does not begin with the outcome. She begins with the structure of the relation. The system is engineered so that the user's next decision is shaped by a designer's model of her weaknesses, and the user is not told this, could not easily discover it, and would likely refuse it if asked directly. The maxim behind the design, roughly "exploit predictable lapses in attention to extend sessions," cannot be universalized without collapse, since it works only while most people do not know it is operating. It also fails the humanity formula, because the user's rational agency is being used as a route to a goal she never adopted. Kant's own example of the false promise makes the parallel clear: the lie succeeds only because its target does not share the liar's knowledge, and the engagement loop succeeds on the same asymmetry.
The harm, on this reading, is not that the user ends up tired. It is that someone decided, without her, how her attention would be spent, and then called the result her choice.
What the Kantian cannot do is rescue the user by appeal to her enjoyment. Kant distrusted happiness as a foundation for morality precisely because it varies from person to person and is contingent on inclination. A feed that delivers pleasure to a user still wrongs her if the pleasure was manufactured by bypassing her judgment. Duty here has less to do with outcomes than with the standing a person retains to set her own ends.
Where the Ledger Looks Healthy
A Benthamite starts from a more uncomfortable position. If a user scrolls for forty minutes and reports, in the moment, that she enjoyed it, the hedonic calculus has a datum. Bentham's own framework lets the individual be the authority on her immediate pleasure, and a utilitarian might argue the platform has simply delivered a product people want. The sum seems to favor the design.
The Benthamite response, though, is not so comfortable as it first appears. Bentham's criteria include certainty and duration, and they require that pains be counted alongside pleasures wherever they fall. The compulsive return, the sleep lost to a late session, the attention drained from work and friendship, and the anxiety of checking are all pains with a long tail. Bentham's list of sanctions, in his chapter on the varieties of pain and pleasure, distinguishes physical, political, moral, religious, and social sources of pleasure and pain, and a feed that colonizes all of them is working on many dials at once. A careful utilitarian has grounds to say that time-on-app is a poor proxy, because it measures how long the hook held, not whether the person was better off.
The Proxy Problem
| Question | Kantian reading | Benthamite reading |
|---|---|---|
| What is the wrong? | Using a person's reason as a means without her adopted consent | Net pain exceeding net pleasure, often unrecorded |
| Does the user's enjoyment settle it? | No, since manufactured pleasure does not license the act | Partly, though momentary reports are weak evidence |
| Is the design wrong if no one is harmed? | Yes, because the wrong lies in the relation | No, if the sum is positive |
| Remedy | Transparency and refusal of covert manipulation | Measure long-run welfare and regulate where the sum is negative |
If you found this valuable, consider supporting our work.
Join PhiloCrux community.
Unlock high-density masterclasses and investigations into ideas surviving outside the algorithmic consensus. Support independent thought and get full access to our digital library.
Join NowWhich Diagnosis Holds the Weight
Shoshana Zuboff's account of behavioral surplus, and Tim Wu's history of attention as a commodity sold to advertisers, both suggest that the harm is structural. The Benthamite can absorb this by treating the platform as an aggregate with a negative externality and proposing regulation. The Kantian can absorb it too, but with a sharper edge: the structure is the wrong, because it is organized around the covert management of persons. The Benthamite fixes the sum. The Kantian indicts the bookkeeper.
The difference matters in practice. A utilitarian reform can be satisfied by a feed that users like better. A Kantian reform requires a feed that does not depend on users failing to understand it. Those are different standards, and a company can meet the first while failing the second.
Referenced Works & Texts
- Kant, Immanuel, Groundwork of the Metaphysics of Morals, Second Section (1785). Formulations of universal law and humanity; the false promise example.
- Bentham, Jeremy, An Introduction to the Principles of Morals and Legislation, Chapter IV (1789). The hedonic calculus and its dimensions.
- Mill, John Stuart, Utilitarianism, Chapter II (1861). Quality of pleasures and the limits of crude summation.
- Eyal, Nir, Hooked: How to Build Habit-Forming Products, Part II (2014). Variable rewards and the engineering of repeated use.
- Wu, Tim, The Attention Merchants, Introduction and Part III (2016). Attention as a commodity sold to advertisers.
- Zuboff, Shoshana, The Age of Surveillance Capitalism, Part I (2019). Behavioral surplus and the structural account of platform harm.