Generative systems do not merely add another kind of media to an already crowded world. They alter the economics of what can be produced, copied, varied, and placed before a human being. A plausible account once required an author, a staff, a printing process, a broadcast slot, or at least the time to compose and circulate it. Those costs never made an account true. They did, however, put some limits on the number of apparently meaningful claims competing for attention.
Synthetic abundance removes many of those limits. A model can produce a story, image, voice, transcript, comment, or apparent eyewitness account at a marginal cost close to zero, then alter its vocabulary for a different audience. The producer can send ten versions where one would once have been possible, and ten thousand where ten would once have exhausted the budget. The resulting problem is not simply that some artifacts are false. It is that the supply of plausible claims can grow faster than the human capacity to notice, rank, remember, and respond to them.
This essay is therefore about information economics and attention governance. It is not an essay about whether a particular photograph is authentic, a question addressed elsewhere in this publication. Its question is what happens when every institution and person receives more plausible material than can be examined, and when the actors who create the flood do not bear the cost of processing it. The scarce resource is not content. It is finite human attention capable of judgment.
The Supply Shock of Plausibility
A synthetic system can generate plausible material without possessing an encounter with the world. It can make a nonexistent meeting sound ordinary, populate a fictional event with incidental details, or produce several explanations for the same rumor. This capacity is useful in fiction, education, translation, accessibility, and simulation. The economic fact remains the same across good and bad uses: the cost of producing a persuasive candidate for attention has fallen dramatically.
Cheap plausibility changes the strategy of manipulation. An attacker need not make one perfect lie and wait for everyone to believe it. The attacker can flood a person, office, platform, or community with enough variants that the recipient must spend time disproving them. The objective may be belief, but it may also be delay, exhaustion, confusion, or the diversion of scarce staff from a real emergency. A claim that is corrected after a deadline, a meeting, or a payment has already done its work.
This is a supply-side attack on institutions. The sender pays for generation; the recipient pays for interpretation. The sender can automate variation, while the recipient must often open each item, identify its relevance, determine whether it duplicates another item, and decide whether a response is required. The asymmetry is especially severe for small parishes, local organizations, schools, families, and public-interest groups that cannot maintain a permanent review staff. Synthetic abundance privatizes the cost of discernment onto whoever is least able to absorb it.
Flooding is successful before it is believed
A flood does not need to produce a coherent alternative reality. It only needs to occupy the queue. Consider an office receiving hundreds of plausible complaints, applications, alerts, or requests in a single afternoon. If every item receives equal initial treatment, the office must either spend its capacity on the flood or allow genuine claims to wait. The attacker has converted a low-cost act of generation into a high-cost allocation problem.
Rate matters as much as volume. Human beings can sometimes examine a large archive over months. They cannot examine an equally large stream arriving before a decision must be made. A rapidly changing stream also prevents stable comparison: a reviewer who finishes one version finds that the claim has already been rewritten, reposted, and surrounded by new variations. In this environment, “we will review everything” is not a commitment. It is a queue with no service capacity.
Attention Is a Governed Common Good
Attention is often described as a private weakness, as if distraction were merely a failure of individual discipline. It is also an institutional resource. A person has limited hours in which to pray, learn, deliberate, care for another, work, and rest. A hospital, newsroom, parish, or civil agency has limited staff time in which to hear a complaint and act on it. When a system deliberately consumes that capacity, it does not merely affect a user experience. It changes whose needs can be recognized.
Catholic anthropology gives this scarcity moral significance. The person is not an infinitely available processor for other people’s demands. Conscience requires the quiet in which a person can attend to truth and responsibility; prudence requires time to judge the concrete good; prayer requires more than an uninterrupted stream of prompts. The Catechism treats prudence as directing judgment toward the true good and meditation as a deliberate exercise of attention. A technology that monetizes the destruction of attention can therefore injure human freedom without stealing a single database record.
Attention governance asks who decides what reaches the front of the queue, what is delayed, what is repeated, and what is made difficult to ignore. That governance already exists in recommendation systems, notification defaults, search rankings, moderation queues, inboxes, and administrative triage. Synthetic abundance makes those decisions more consequential. If a platform treats every generated item as a fresh opportunity for engagement, it gives the producer of the flood power over the time of everyone downstream.
Platform Incentives and the Commercial Flood
Many platforms are not paid simply for storing information. They are paid for visits, sessions, impressions, subscriptions, conversions, or data about attention. Their ranking systems therefore have an incentive to prefer material that provokes a response quickly. Novelty, anger, fear, intimacy, and apparent certainty can outperform patient qualification. A synthetic system can supply these traits at scale, testing thousands of headlines, thumbnails, personas, and emotional framings against the platform’s audience.
The platform need not intend deception for its incentives to reward the flood. A ranking model that measures immediate interaction can promote a high-volume producer over a careful source whose work takes time to read. A creator program can pay for output rather than usefulness. A notification system can turn every generated reply into an urgent-looking demand. The result is a market in which the costs of production fall while the costs of attention remain human.
“More choice” is an incomplete defense. Choice helps when people can compare options under conditions of adequate time and intelligible difference. A thousand near-duplicate claims do not necessarily expand agency. They can make refusal, comparison, and memory more expensive. A person who cannot tell which items deserve a first look is not sovereign over an abundant market. The person is being governed by the order in which the market presents its demands.
Search and Archive Contamination
Synthetic abundance also changes the temporal structure of knowledge. Search indexes, public archives, document repositories, and retrieval systems are built from accumulated traces. A fabricated account can be copied into many pages, quoted by later summaries, translated, embedded in a discussion, and eventually retrieved as though its recurrence demonstrated importance. The original generation event disappears while the derivatives remain available to search.
This is not just a problem of false answers. It is contamination of the material from which future questions are answered. A researcher may encounter ten references that all descend from one generated paragraph. A retrieval-augmented system may select a fluent derivative because it matches the query, then produce another derivative that enters the archive. Each step can increase textual presence without adding a new contact with reality. The archive becomes large while its independent informational value shrinks.
Corrections face an unfavorable economics. The original claim may have been short, vivid, and easy to copy; the correction may require context, qualifications, and links to primary material. Search and memory often preserve the concise claim more readily than the careful repair. A responsible archive therefore needs more than deletion. It needs relation: which records repeat one another, which claims are disputed, what was generated or transformed, and which later account corrects an earlier one. Without such context, abundance creates a false impression of historical depth.
From Verification to Queue Design
The answer cannot be to verify every item with equal intensity. That promise ignores finite capacity and rewards whoever can produce the most work for others to inspect. Institutions need explicit queue design. They should distinguish urgency from volume, consequence from vividness, and reversibility from novelty. A claim that could trigger immediate harm deserves rapid attention; a low-stakes duplicate can wait or be grouped with related material.
Such triage should be provenance-neutral. This does not mean provenance is irrelevant when a particular claim is examined. It means that the presence of a credential, institutional origin, or polished identity should not be the sole ticket to consideration, while an uncredentialed source is not automatically sent to oblivion. Prioritization should ask about stakes, vulnerability, time sensitivity, independent routes to the claim, and the cost of being wrong. It should not turn a technical badge into a hierarchy of whose experience is admissible.
Rate limits are part of this moral architecture. They can restrict mass creation of accounts, repeated mentions, automated replies, bulk notifications, or rapid submission of near-duplicate material. Cooling periods can give a person time to reconsider an emotionally charged action. Batching can prevent one producer from occupying every review slot. These measures are not assertions that the delayed material is false. They are assertions that no actor is entitled to consume an unlimited share of a finite community’s attention.
Queues must remain contestable. A system that silently deprioritizes claims can hide bias behind efficiency. People need to know whether their request is delayed because it is low stakes, duplicated, incomplete, or outside the institution’s purpose. They need an escalation path for imminent danger and a way to correct an automated grouping error. Good governance does not promise everyone immediate attention. It makes the allocation of attention intelligible and revisable.
Limits and Tests for Triage
“Infinite” should be read as a rate-and-capacity mismatch, not as a claim that production is literally without limit. The relevant comparison is among the arrival rate of material, the institution’s service capacity, and the time before a decision matters. Generation, distribution, and intake are different points in that system. A platform may limit automated distribution without limiting speech; an office may batch intake without accepting that a claim is unimportant. The essay does not have contemporary measurements showing how often synthetic derivatives contaminate a particular archive. That is an evidence gap, not permission to treat every prolific speaker as an attacker.
Grouping must also leave room for independent witnesses. Similar wording is not proof of common origin: several people may describe the same neglected harm, especially when they share a place or experience. A triage system should preserve each source’s route to knowledge, flag uncertainty about derivation, and escalate a cluster when its accounts differ in relevant details or when a vulnerable person reports that delay creates danger. An appeal log should record false grouping, missed escalation, and the time needed to correct both. If a recipient repeatedly refuses to process material despite credible urgency or available capacity, that refusal is itself a governance decision rather than a neutral consequence of abundance.
These proposals need testing rather than confident prediction. A newsroom, parish, or public agency could compare a queue before and after batching or rate limits, measuring time to a genuine claim, false-positive grouping, escalation of vulnerable sources, abandonment, and correction reach. Independent review should examine whether uncredentialed or minority speakers are delayed disproportionately. Until such pilots exist, Simon’s account of attention scarcity and Wu’s history of commercial attention frame the problem; they do not establish the size of a generative-media effect or validate a particular ranking rule.
Strong Objections
“Synthetic abundance democratizes expression”
It can. Translation, drafting, captioning, and inexpensive production can let a small community speak where cost once excluded it. A campaign for a neglected cause may need many local versions of a message. Any response that treats volume itself as suspicious would preserve the advantage of institutions that already possess staff and distribution.
The distinction is between access to expression and a right to occupy every listener’s queue. Democratic participation requires channels for new and unpopular speech. It does not require platforms to deliver unlimited duplicates at maximum speed, or institutions to treat a mass-produced demand as thousands of independent requests. Rate, grouping, and transparent triage can preserve expression while preventing volume from masquerading as public importance.
“Attention filters will become censorship”
They can become censorship if private or public authorities use them to suppress disfavored viewpoints, conceal misconduct, or make appeal impossible. The answer is not to abandon filtering; every finite person and institution already filters. The answer is to constrain the filter’s purpose, publish meaningful rules, preserve avenues for challenge, and distinguish withholding recommendation from forbidding speech. A queue policy should reduce extraction of attention, not decide which convictions may exist.
“Slowing the stream protects the powerful”
Delay can indeed favor incumbents when the powerful control access to records or can wait out public interest. A poor queue design treats urgency as volume and asks a vulnerable person to submit the right form before receiving protection. Triage must therefore reserve expedited paths for credible risk, retaliation, disappearing evidence, and people with limited ability to navigate the system. Protecting attention is not permission to make the vulnerable wait quietly.
Practical Implications
Platforms should measure the external cost of generated volume, not only the engagement it produces. Useful controls include caps on automated outreach, deduplication of near-identical submissions, user-controlled notification budgets, friction before mass forwarding, and clear distinctions between an item being stored, recommended, and urgently escalated. Ranking systems should reward durable usefulness and responsible context rather than treating every reaction as proof that more exposure is deserved.
Archives and search systems should preserve relationships among records. They can cluster derivative copies, mark unresolved claims, retain correction histories, and show when many references descend from one source. Retrieval systems should be tested for synthetic feedback loops in which generated summaries become evidence for later summaries. The aim is not to purge every generated sentence. It is to stop textual multiplication from being mistaken for independent knowledge.
Institutions should publish service rules for attention. A parish office can state which messages receive immediate pastoral escalation, which requests belong in a normal queue, and how an urgent concern can bypass routine delay. A public agency can separate mass duplicates from distinct claims without pretending the duplicates are irrelevant. A school or nonprofit can set limits on automated submissions while maintaining a human route for safeguarding concerns. These are administrative details with moral consequences.
Individuals also need protected intervals in which no platform is entitled to compete for attention. Turning off nonessential notifications, refusing instant replies, and choosing a small number of deliberate sources are not retreats from public life. They are ways of recovering the silence required to decide what deserves response. Communities can support this discipline by not treating constant availability as charity and not confusing rapid reaction with solidarity.
Sources and qualifications
The account of finite attention and information abundance draws on Herbert A. Simon’s “Designing Organizations for an Information-Rich World” (1971), whose economic formulation treats attention as the scarce resource when information is abundant, and on Tim Wu, The Attention Merchants (2016), concerning the commercial capture of human attention. These works do not provide a complete theory of generative systems or a ready-made platform policy. They clarify the economic problem this essay applies to synthetic production and distribution.
The Catholic account draws on the Catechism of the Catholic Church 1776, 1806, 2705–2708, and 2730; Gaudium et Spes16 and 26; and Inter Mirifica 5–8. These sources do not name attention governance as a technical field. They support the claims that conscience, prudence, contemplation, dignity, and participation are human goods, and that communications should be ordered toward the person and the common good rather than treating people as available inventory for extraction.
“Flooding,” “queue,” “rate,” and “contamination” are used here as economic and institutional descriptions, not as claims that every generated item is deceptive or that every high-volume speaker is acting maliciously. Queueing capacity, ranking methods, archive design, and legal duties vary by institution and jurisdiction. Provenance-neutral prioritization does not prohibit provenance checks when a claim is examined; it rejects provenance status as the sole determinant of whose claim receives attention. This essay is not legal, technical, or pastoral advice.
Protecting Attention for Reality
Infinite testimony is not dangerous only because some testimony is false. It is dangerous because an unlimited supply of plausible material can consume the finite time by which people distinguish urgency from noise, a new claim from a copy, and a serious need from a demand engineered to look serious. The decisive contest is over the queue: who can fill it, who must work through it, and whose claim is left waiting when capacity is gone.
A humane response will not promise perfect attention or require every citizen to investigate everything. It will govern scarcity openly. It will limit automated extraction, preserve routes for urgent and vulnerable claims, group duplication without erasing difference, protect archives from synthetic self-repetition, and make triage answerable to the people it delays. It will treat friction not as a defect in every circumstance but as a safeguard against preventable consumption.
Catholic technology should begin here: the human person is not a vessel into which an unlimited market may pour its demands. Attention is part of the person’s capacity to know, love, deliberate, pray, and act. When systems protect that capacity, they do more than reduce spam. They make room for reality to be encountered at human speed, for responsibility to be assigned, and for the claims that matter to receive something synthetic abundance cannot manufacture: a finite person’s careful attention.
Source Notes
Primary text: §1776 on conscience, §1806 on prudence, §§2705–2708 on meditation, and §2730 on vigilance and attention in prayer.
Primary text: §§16 and 26 on conscience, human dignity, truthful social life, and the common good.
Primary text: §§5–8 on communications media, truthful representation, human dignity, and the responsibilities of creators and users.
Bibliographic record for the 1971 edited volume Computers, Communications, and the Public Interest; Simon’s chapter frames attention as the scarce resource in information-rich environments.
Publisher page for Wu’s 2016 book on the historical development of attention markets and the commercialization of human attention; a secondary historical account, not a primary source.
Editorial revision note
This essay completed one self-editing, cross-essay consistency, and strongest-criticism pass. Its controlling thesis was preserved while objections, qualifications, evidence limits, and practical consequences were strengthened.