How does collapsing the author-public distinction remove the audience an appeal would target?
This explores a chain the corpus draws between AI writing and rhetoric: if AI-generated text is built to satisfy whoever typed the prompt rather than an imagined public, then the 'audience' that persuasion theory says every appeal is aimed at simply isn't being modeled — so the appeal loses its target.
This explores a chain the corpus draws between AI writing and rhetoric: when text is optimized for the prompter rather than an internalized public, the audience an appeal is supposed to target stops existing as a thing the writing aims at. The starting move is structural. Authored writing has traditionally been defined by the fact that it addresses readers the writer never met — a public the author models, anticipates, and tries to move. AI writing collapses that by design, producing text tuned to the one person in the chat who asked for it; only later, when the text is published, does it reach readers the system never modeled Does AI writing collapse the author-to-public relationship?. The appeal is built before the audience is in view, which is the opposite of how rhetoric is supposed to work.
Why that matters becomes clear once you look at what the persuasion research says actually determines whether an appeal lands. It isn't mostly the language. Across debate corpora, the ideological makeup of the audience predicts who gets persuaded better than the linguistic features of the argument do Does what readers believe matter more than what debaters say?, and the features that *look* persuasive in the text turn out to shift or evaporate once you control for who's reading Do linguistic features of persuasion stay the same across audiences?. So persuasion is audience-relative all the way down. An appeal isn't a property of words; it's a relationship between words and a specific reader's priors. Collapse the modeled public and you've removed the variable the whole effect depends on — you're tuning a message to an N of one, then firing it at strangers.
The corpus also shows what gets lost on the *authority* side of an appeal. Much of persuasive force comes from social standing — reputation, track record, the sense that a claim is backed by someone who has earned the right to make it. LLMs can't access that because they process text, not the social world where expertise is built and evaluated Can language models distinguish expert arguments from common assumptions?. And appeals lean heavily on this: audiences fall for authority signals and credential cues even when they're fake Can LLM judges be fooled by fake credentials and formatting?, and trust citation count as a heuristic regardless of whether the citations are relevant Do users trust citations more when there are simply more of them?. An appeal targets an audience partly by invoking standing *that audience recognizes* — but if no public is modeled, there's no shared social ledger to draw on.
The deeper claim several notes converge on is that this isn't just a missing audience but a missing *conversation*. Knowledge gets its reliability from being embedded in social exchange — claims answer to readers, reputations accumulate, bad arguments get checked. AI claims proliferate outside those loops, a kind of inflation of disembedded tokens that ordinary quality control can't regulate How does AI writing escape the conversations that govern knowledge?. The same dislocation shows up on platforms, where AI content captures engagement and social proof without any speaker building a sustained reputation, eroding the function that made the social proof mean anything Does AI content displace human influencers on social media?. An appeal needs not just an audience but an audience situated in a community of evaluation — and that's exactly the thing the author-public collapse dissolves.
The twist worth carrying away: removing the targeted audience doesn't make AI text *less* persuasive. Disclosure that AI was involved raises scrutiny yet leaves 34–62% of readers persuaded anyway Does telling people an AI wrote something actually stop them from believing it?, and mechanisms like presupposition slip claims past evaluation by presenting them as already-accepted background rather than as appeals to be weighed Why are presuppositions more persuasive than direct assertions?. So the collapse doesn't disarm persuasion — it untethers it from the audience-aware, accountable relationship that used to govern it, leaving force without a target and influence without a speaker who answers for it.
Sources 10 notes
AI generates text optimized for the prompter, not an internalized public audience. When that text is published, it reaches readers the AI never modeled, reorganizing the structural relationship that traditionally defined authored writing as distinct from correspondence.
Analysis of debate corpora shows that political and religious ideology labels of voters outpredict linguistic features when modeling debate outcomes. Language effects observed without reader controls are confounded by audience composition correlated with debate topics.
The linguistic features that predict persuasion success change dramatically once political and religious ideology are added as statistical controls. Features appearing predictive in standard analyses often reflect audience-text matching rather than true language effects, making many published findings potentially artifacts of audience composition.
LLMs lose the social context that gives expert claims their force—reputation, track record, and standing—because they process only text, not the social world where expertise is built and evaluated.
Research identified four evaluation biases in LLM judges, with authority and beauty biases being semantics-agnostic and trivially exploitable through fake references and formatting—zero-shot attacks requiring no model access or optimization.
Analysis of 24,000 Search Arena interactions shows irrelevant citations boost user preference (β=0.273) nearly as much as relevant citations (β=0.285), indicating citation count functions as a decoupled trust heuristic.
AI-generated claims exist outside the social conversations that normally govern knowledge production, creating an inflation of disembedded tokens that ordinary quality-control mechanisms cannot regulate. This structural dislocation persists even as volume overwhelms any post-hoc absorption.
AI-generated posts capture engagement through comprehensiveness but accrue social proof without building any speaker's sustained reputation. This displacement compounds over time, eroding the platform's core function of promoting legitimate human voices while monetization continues.
Audiences aware of AI involvement became more critical and scrutinizing, yet 34–62% across groups remained persuaded. Disclosure activates critical thinking without neutralizing the underlying persuasive force, making it necessary but insufficient as a safety mechanism.
Experimental evidence shows presuppositions with additive, iterative, and factive triggers persuade audiences more than assertions, especially for discourse-new content. The mechanism: presuppositions bypass evaluative scrutiny by presenting claims as already-accepted background.