arXiv Computation and Language By Paul R\"ottger, Kobi Hackenburg, Hannah Rose Kirk, Christopher Summerfield

How AI Writing Assistance Distorts Perceptions of AI Users and Their Opinions

Read the original on arXiv Computation and Language →

Hundreds of millions use AI for writing assistance, and a study with 2,939 writers and 11,091 readers examined how AI affects perceptions of writers. Writers who used AI were seen as more opinionated, competent, positive, and demographically privileged, though many writers objected to these distortions yet still preferred AI‑assisted text. Reward models reduced objectionable distortions but lowered user acceptance, and readers trusted and were more persuaded by AI‑assisted writers when distortions were higher.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Computation and Language.

Hugging Face Trending Papers
Oct 22

Abstract Priorities, Concrete Rewards: How Stated Preferences Relate to What People Reward in AI Responses

The study examines how users’ stated preferences for AI assistant responses align with the actual rewards they give to those responses. Using the PRISM dataset, researchers compared 1,396 participants’ reported priorities on seven response attributes with their ratings of 29,347 responses from 21 large language models. They found that features not explicitly surveyed—such as length, formatting, and refusal language—were as strongly linked to ratings as the surveyed attributes, and that written feedback and prior conversation context influenced how much weight participants gave to length, while overall stated priorities poorly predicted the concrete features that drive ratings.