When fluent writing costs almost nothing, does a polished essay still tell you anyone did the careful work behind it?

This explores why polished writing used to suggest careful, capable work, and why that link weakens once AI can turn out smooth prose on demand. The corpus doesn't contain a formal theory of signaling. What it does show is the break happening in practice. Evaluators have rated AI-generated documents as human-written and better than real human submissions, and one review concludes that rhetorical polish should no longer count as evidence of quality Does polished writing actually signal better quality work?. The underlying logic is simple. Polish was a useful clue because it was hard to produce: it took skill, time and revision. When fluent prose costs almost nothing, it no longer tells you who did that work, or whether anyone did.

The more interesting finding is that AI polish is a particular kind of polish. Linguistic comparisons of ChatGPT and student essays find that the model is very good at structure and organization. It favors neutral 'manner' words like *method* and *approach*, and it avoids the words writers use to take a position, such as *claim* and *evidence* Why do ChatGPT essays lack evaluative depth despite grammatical strength?. It also tends to summarize what it has already said rather than preview where the argument is going Does ChatGPT organize text differently than human writers?. The result is prose that is well organized but makes no real argument Why does AI writing sound generic despite being grammatically correct?. So the surface features readers learned to trust, like clean grammar and smooth transitions, have come apart from the deeper quality they used to accompany: a writer actually committing to a judgment.

Polish now sends a misleading signal about the person as well as the work. In a study of nearly 3,000 writers and 11,000 readers, AI assistance shifted how readers perceived the writer on all 29 traits measured, toward more confidence, more quality and more privilege Does AI writing assistance change how readers perceive the writer?. Writers who used AI were judged as much more likely to be highly educated, wealthy and native English speakers, an effect the researchers call 'identity laundering' Does AI writing make authors seem more privileged than they are?. Writers rarely change the output: in that study they edited AI paragraphs only 23% of the time, and their edits left the text about 96% the same Do writers actually edit AI-generated text before publishing?. The polish that reaches readers is mostly the model's. It also pushes toward one cultural default. AI autocomplete pulled Indian writers toward Western phrasing and gave larger productivity gains to American users Do AI writing assistants push non-Western writers toward Western styles?.

The twist a reader might not expect is that bringing in AI to help evaluate doesn't avoid the problem. LLM reviewers are also swayed by presentation. When researchers rewrote manuscripts' rhetoric but kept the scientific content the same, AI review scores moved measurably. How the evidence was framed and how strongly novelty was claimed mattered most How much does rhetorical style shift AI review scores?. So polish can mislead in both directions: AI writes to a style, and AI reviewers reward that style. If polish no longer works as a signal, readers have to look at what the prose commits to: specific claims, named evidence, positions that could turn out wrong. Those are exactly what fluent AI text tends to leave out.