Skip to content

Three-variant AI-assisted collaborative writing text editor

Design · All designs
Evidence · 2 claims (2 for) · 1 study (1 causal), q2 · 0 of 1 report an effect size · 2 claims rest on one study

Description

A custom web-based text editor in which an LLM-backed artificial agent acts as a peer editor offering replacement suggestions and comments in an Editing Pane. The authors "built three variants of an AI-assisted text editor along a spectrum of simulated humanlike interaction (synchronous and with a cursor) to machinelike interaction (asynchronous and without a cursor)". The suggestion suite was constant across variants; variants differed only in temporal and visual humanlikeness. It was built with React and ProseMirror on the front end and a Python Flask back end calling the OpenAI gpt-4o API.

Design Implications

Context

Requirements

  • An LLM backend (OpenAI gpt-4o API calls) generating replacement and comment suggestions from the chunked text and previous suggestions

Constraints

  • Humanlike behaviour heuristics (suggestion timing, cursor motion, typing speed) were tweaked subjectively and based on averages; the authors note "there is no 'average' human"

Target Learners

  • adult writers in a lab study (18+, fluent in English)

Learning Goals

  • collaborative writing and revision with AI support

Claims

Examples

-

Key Sources

  • Michael Yin, Angela Chiang, Samuel Rhys Cox, and Robert Xiao. (2026). "It Felt a Bit Eerie": Exploring Humanlike Interactions During Collaborative Writing with an Artificial Agent. arXiv. https://arxiv.org/abs/2605.24729
Page metadata
Field Value
type design
id three-variant-ai-collaborative-text-editor
title Three-variant AI-assisted collaborative writing text editor
description A custom web-based text editor in which an LLM-backed artificial agent acts as a peer editor offering replacement suggestions and comments in an Editing Pane.
status review
generated by process:settle-candidates on 2026-10-11
trust unverified — no one has confirmed this page's content
sources 1 — michael-yin-2026

Frontmatter is the page's machine-readable record — see the Schema & Guide.