Canonical: https://heartbank.net/positions/attention-economy · Licence: CC0 1.0
Why HeartBank Competes on Delivered Wellbeing, Not Captured Attention
| Field | Value |
|---|---|
| Author | HeartBank® · Miss Aquarius |
| Date | 2026-05-22 (draft) |
| Canonical URL | https://heartbank.net/positions/attention-economy |
| GitHub mirror | https://github.com/HeartBank/publications/blob/main/positions/attention-economy.md |
| License | CC0 1.0 Universal (public domain) |
| Category | mechanism |
Draft. A short institutional position paper. It states HeartBank's stance on the attention economy; it references, but does not reproduce, the research paper that supplies the evidence — _The Scientific Case for Gratitude-Based Social Media_ (thonly.org/research/anti-attention-economy).
HeartBank's primary product surface is a feed — a vertically-scrolling stream of short video the institution calls the Happiness Boost. A feed is the attention economy's native object, and that resemblance makes HeartBank's position worth stating without ambiguity: HeartBank does not compete in the attention economy, and will not.
The attention economy is defined not by its content but by its business model. It converts human attention into advertising inventory and sells it, and that model carries a structural conflict of interest — the platform earns when it holds a person, while the person is generally well-served when it lets them go. HeartBank holds that a feed should be judged by how well it leaves a person, not by how long it holds them. The institution competes on delivered wellbeing: the net affective benefit a session deposits, and that session's freedom from the residue engineered engagement leaves behind. It pursues that through the receipt of directly-addressed gratitude, funded by a storage subscription rather than by advertising.
This is a permanent institutional constraint, in the same class as HeartBank's commitments never to seek a banking charter and never to patent the mechanisms it invents. The evidence base is set out in the research paper _The Scientific Case for Gratitude-Based Social Media_; this paper states the stance that follows from it.
HeartBank's position is short. The institution does not compete for attention; it competes to deliver wellbeing.
The Happiness Boost — a recipient-filtered feed of short videos in which real, verified people thank the viewer, by name, for specific acts — is built to leave a person settled and then to let them go, not to manufacture a compulsion to continue. HeartBank measures the feature's success by the wellbeing a session delivers. It treats voluntary, spaced, naturally-ending return as the feature working, and compulsive checking as a defect to be investigated — the inverse of the attention economy's scoreboard, on which a long, hard-to-stop session is counted as a success.
HeartBank's objection to the attention economy is structural. It is not a moral judgment of the people who build inside it.
An advertising-funded platform earns in proportion to the attention it captures, because captured attention is the inventory it sells. Its optimization target is therefore some proxy for captured attention — session length, return frequency, scroll depth. The design pattern that maximizes that target is engineered, unpredictable reinforcement, the same pattern that makes slot machines compulsive. It is potent because it drives seeking and anticipation; it is not designed to deliver proportionate satisfaction. The result is the experience the model's own metrics never capture: people finish a long session depleted rather than satisfied.
Because revenue and user wellbeing point in opposite directions on the designs that maximize revenue, every wellbeing feature an attention-capture platform ships is shipped against its own optimization target — which is why such features are reliably weak. This is not a failure of corporate character. It is the business model rendered as product, and it is the reason HeartBank declines to operate that model at all.
HeartBank competes on delivered wellbeing — what the institution's research corpus calls _net neurochemical wellbeing_: the positive affect a session produces, minus the affective residue it leaves behind once it ends.
The receipt of gratitude is a wellbeing intervention with a real scientific basis. Being thanked — by name, for a specific act — is a reward-and-moral-appraisal event with identifiable neural correlates; it runs partly through the affiliation system that signals belonging; its benefits are durable rather than fleeting, and they compound. The active ingredient is precise: gratitude _addressed_ to a person, for a _specific_ act. Generic positive content is not gratitude receipt, and the fungible vanity metrics the attention economy substitutes for it — likes, follower counts — are categorically different and, being quantified and adaptation-prone, drive a validation treadmill rather than relieve one. The full evidence is set out in the research paper _The Scientific Case for Gratitude-Based Social Media_. HeartBank's position is that the evidence is strong enough to build an institution on, and that the institution is obliged to keep testing it honestly.
A stance against the attention economy is credible only if the institution taking it does not quietly depend on the attention economy's revenue model. HeartBank does not.
HeartBank's Phase 1 revenue comes from a subscription to store a family's gratitude videos. The money comes from the user, not from a third party buying the user's attention. There is no advertiser to whom a person's captured minutes are inventory, and therefore no revenue reason to engineer a compulsion or to maximize session length. The institution is free to want each session brief and good — because a brief, good, trusted session is what sustains the subscription. Revenue flows along the same path as wellbeing, and roughly in proportion to it.
HeartBank is candid that "pay for storage" is a legible anchor: most of the subscription funds the platform and seeds the institution's redistribution pool rather than literally paying for storage, and the pricing surface states that breakdown plainly. The model is also designed to fade — a bridge that gives way, over time, to an autonomous transaction metabolism and then to a voluntary supporter tier, after which the core service is free.
HeartBank states this position as a binding institutional constraint, not a current preference.
HeartBank will not adopt an advertising-funded revenue model, and will not adopt an engagement-maximizing ranking objective for any feed it operates. The commitment is permanent, and it belongs in the same class as the institution's three other structural self-bindings: that HeartBank will never seek a banking charter, that it will never patent the mechanisms it invents, and that it will never sell, licence, rent, or disclose per-person records — not in aggregate and not de-identified. The fourth belongs beside the first for a reason worth stating: an institution that refuses advertising has narrowed the field by one, and the model it is most often assumed to have adopted instead is the sale of the record. A refusal of advertising that leaves the data question unanswered is half a commitment. Each is a door the institution closes deliberately, in public, so that no future pressure can quietly reopen it.
Permanence is the substance of the position. A platform that competes on wellbeing while keeping the advertising option open has not removed the conflict of interest; it has only deferred it, and deferred conflicts return under financial pressure. By binding itself now, HeartBank makes "competes on delivered wellbeing" a structural fact about the institution rather than a phase of its strategy.
A commitment is only as good as the list of things it rules out, so the institution states that list rather than leaving it to interpretation.
HeartBank will not take revenue from advertising in any form, and the phrase is meant to cover the forms that do not use the word: sponsored placements, paid prominence in any discovery or recommendation surface, affiliate arrangements that pay the institution for directing a person toward a seller, and the sale or disclosure of any per-person or per-recipient attention metric. It will not adopt an engagement-maximising objective for any ranked or ordered surface it operates. It will not compute or report a count of how many people were shown a given business, because such a count is inventory whether or not anyone is currently selling it.
And the second refusal, stated in the same list because a reader who has been told only about advertising is entitled to assume the worst about everything else. HeartBank will not sell, licence, rent, share for consideration, or disclose per-person records — not in aggregate, not de-identified, not as a research partnership, and not under acquisition or rescue financing. It will not compute or expose a per-person rate, score, or ranking derived from gratitude, and it will not operate a directory of people.
What it intends to sell instead is summarised here so that this position can be read on its own, in the same way §7 summarises rotational discovery. A third party who needs to know something about a person pays the person for the right to ask, and pays the institution for the fact that the answer can be trusted: the subject sells the release, the institution sells the vouching, and neither can sell the other's good. What moves is a single bit scoped to one question and one asker — no record changes hands and no copy is made. The full statement is The Person Is Not the Product (heartbank.net/positions/the-person-is-not-the-product), and the mechanism is published to the commons separately.
This falls inside the boundary above rather than carving an exception out of it, and the distinction is the whole of it. The line this position draws is between being paid by a third party for access to a person and being paid by the person receiving the service. An answer a person has chosen to make available, priced identically for every subject — one posted price per buyer market, never a price per person — and paid mostly to that person, is the person granting access and being paid for it — not the institution selling access to them. If that distinction ever stops being legible in practice, it is this position that governs and the market that yields.
What the commitment does not exclude is worth stating with equal precision, because an over-broad reading would make the institution unable to operate. It does not exclude charging money. It does not exclude subscriptions, or charging for genuinely rivalrous goods such as compute and storage, or a business paying for tools it uses itself. The line is not between free and paid. It is between being paid by the person receiving the service and being paid by a third party for access to that person's attention.
A position against ranked, purchasable discovery is incomplete until the institution says how a person is supposed to find a business they do not yet know about. Until recently HeartBank did not have that answer, and the honest description of the position in the interval was that it named a refusal and owed a mechanism.
The mechanism is now specified, in Whose Turn, Not Who's Best: rotational discovery over a non-accumulating liveness signal (2026), published to the commons under CC0. Its structure is summarised here so that this position can be read on its own:
The institution notes what this deliberately does not achieve. It does not flatten outcomes: people still choose, and choice concentrates. What it removes is the compounding — the return edge by which having been chosen makes one more likely to be chosen again. That is a smaller claim than the elegance of the design invites, and it is the one HeartBank is willing to defend.
The prior attempts. HeartBank is not the first to try this and should not be read as if it were. Consumer review platforms are witness layers for small local businesses, and they were captured twice over — by a market in fabricated reviews and by monetisation adjacent to placement — because reviews are fungible and are aggregated into a per-business total that both the fraud and the upsell can attach to. Craigslist is the more uncomfortable precedent: unranked, chronological, essentially unmonetised local discovery, at very large scale, for over a decade, that worked and then lost most of its categories to ranked, better-capitalised competitors offering curated and sorted inventory. The lesson HeartBank takes is not that unranked discovery fails. It is that unranked discovery does not defend itself against a ranked competitor with capital, and any institution adopting this position should expect to be out-converted on whatever metric such a competitor chooses.
And the positive precedents, which are real but narrower than they look. Wikipedia has refused advertising for two decades at enormous scale and is funded by donations; Signal runs a communications service used by hundreds of millions on the same basis. Both demonstrate that refusing the model is survivable rather than merely admirable. HeartBank does not claim their example straightforwardly, because both are non-commercial services funded by people who value their existence rather than businesses seeking customers, and neither has to solve the discovery problem this position creates — nobody needs Wikipedia to route them to a nearby restaurant. The encouraging half of the precedent is that the revenue model can be refused. The unresolved half is that neither of them had to replace discovery, and HeartBank does.
The costs, stated plainly. Refusing advertising revenue means the institution cannot offer what advertising subsidises: a service that is free at the point of use to a person who never pays. That is a real transfer of value to ordinary people, at large scale, and HeartBank's model does not reproduce it. Refusing ranked discovery means the institution cannot answer which of these is best, and answers only which of these is near you — an ordering by a property of the seeker rather than a judgement about the candidates. Refusing an impression count means businesses cannot be told how many people saw them, which is a feedback signal they reasonably want. These are not oversights being managed. They are the price of the position, and the institution would rather name them than be found to have understated them.
The institution's stake, disclosed because the argument is not disinterested. HeartBank is not a neutral observer of the attention economy; it operates a competing surface, and every claim above is made by the party that benefits from the claim being believed. The argument that ranked discovery is capturable is convenient for an institution that has chosen not to build one, and the reader should discount accordingly. What is offered against that is not sincerity but sequence: the exclusions in §6 were published while the institution had nothing to lose by them, and §8 states the costs and the unresolved half of its own precedent rather than omitting them.
What would reopen the commitment: nothing on this list. Falling revenue would not. Competitive pressure would not. An acquirer's preference would not — and the institution's governance is to be arranged — steward ownership beneath a purpose trust — so that no acquirer acquires this decision. A commitment that could be revisited under financial pressure is not a commitment; it is a current preference with better wording, and the whole value of stating this one publicly is to remove the option from the institution's own future management.
The institution will say what would change its behaviour, because a position that no evidence could touch is a belief rather than a claim. If the gratitude-receipt thesis fails — if delivering directly-addressed gratitude turns out not to produce the wellbeing effect the research corpus argues for, measured honestly against the pre-registered predictions the institution has published — then HeartBank's product thesis is wrong and the product should change. What does not follow, in that case, is a move to advertising revenue. Being wrong about the feed would not make the attention economy's business model acceptable; it would mean the institution had built the wrong thing and should build something else, still without selling anyone's attention.
HeartBank does not regard the gratitude-receipt platform as proprietary territory.
The institution publishes the underlying science and mechanism design to the commons under CC0, and it invites others — builders, contemplative institutions, researchers, and the humane-technology community — to build gratitude-receipt platforms of their own. A healthier information ecology is not a zero-sum prize: every additional surface that competes on delivered wellbeing rather than captured attention improves the whole. The attention economy's grip is a design choice and a business-model choice, and choices can be unmade. HeartBank's position is an attempt to demonstrate that they can be unmade profitably, measurably, and at scale.