AI as an Active Writer - Interaction strategies with generated text in human-AI collaborative fiction writing

Aug 28, 2026

Machine Learning (ML) supports human fiction writers by extending inspiration sources and creative strategies. A web-based human-AI collaborative writing tool was prototyped to help writers shorten, edit, summarize, and regenerate AI-produced text. Using a “finish each other’s story” turn-taking approach, a preliminary study with 9 users analyzed interaction dynamics. Results show users valued unexpected machine text, expected reduced fluency and coherence when editing was enabled, and formed a mental model of the AI as an active writer rather than a passive assistant, informing co-creative design implications.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages10

Reading Time25 min

AccessFree to download

Abstract

1. Introduction

Human-AI collaborative creative systems

User needs in creative writing

Method and Prototype Design

Web-based collaborative writing tool

Two interaction modes

Study procedure and GPT-2 generation

Findings and Discussion

Patterns in collaboratively written stories

User perceptions of AI fluency and coherence

AI as an Active Writer: Interaction strategies with generated text in human-AI collaborative fiction writing Daijin Yang1 , Yanpeng Zhou2, Zhiyuan Zhang3 , Toby Jia-Jun Li4 and Ray LC3 1 Northeastern University, 360 Huntington Ave, Boston, MA 02115, USA 2 Nanyang Technological University, 50 Nanyang Ave, 639798, Singapore 3 City University of Hong Kong,Tat Chee Avenue, Kowloon Tong, 518057, Hong Kong 4 University of Notre Dame, Notre Dame, IN 46556, USA Abstract Machine Learning (ML) has become an important part of the creative process for human fiction writers, allowing them to utilize various sources of information and be inspired by strategies and data previously seldom explored. To investigate how human writers collaborate with ML systems in fiction writing, we prototyped a web-based human-AI collaborative writing tool that allows writers to shorten, edit, summarize, and regenerate text produced by AI. To investigate the dynamics of human-AI interaction in fiction co-writing, we used a "finish each other’s story" approach where humans and machines took turns writing collaborative fiction. In results from a preliminary study with 9 users, we found that users took inspiration from unexpected text generated by the machine, that users expected reduced fluency and coherence in the machine text when allowed to edit the output, and that they perceived a mental model of the AI as an active writer in the collaborative process rather than simply as a passive AI writing assistant. This study provides design implications on supporting co-creative writing of humans and machines. Keywords Applications of intelligent user interfaces, Collaborative interfaces, User Modelling for Intelligent Interfaces, Evaluations of intelligent user interfaces-Reproducibility 1. Introduction The rapid development of machine learning has made it possible for artificial intelligence (AI) to collaborate with humans to generate creative content [1, 2, 3, 4, 5, 6] . Human-AI collaborative creative systems based on machine learning have been gradually entering people’s creative artistic life such as music composition [6, 7, 8], creative illustration [1, 9], and co-writing [10, 11] . These human-AI collaborative creation systems can assist experienced creators by inspiring them with new ideas and providing suggestions [12, 13] . They can also bring a novel creative experience to users who have no or little creative experience, such as completing the drawing that the user has started or automatically filling in the user’sunfinished sentence [1, 10] . In this article, we focus on the needs of users when they collaborate with AI for creative writing. Recent work is focused on improving the algorithmic performance of natural language generation models, such as improving the logic of generated text [14, 15] or making the generated text closer to the natural language [16, 17] . However, little work focuses on explor- Joint Proceedings of the ACM IUI Workshops 2022, March 2022, Helsinki, Finland $ [yang.dai@northeastern.edu](yang.dai@northeastern.edu) (D. Yang); [zydstd@gmail.com](zydstd@gmail.com)[ ](zydstd@gmail.com)(Y. Zhou); [zzhang452-c@my.cityu.edu.hk](zzhang452-c@my.cityu.edu.hk) (Z. Zhang); [toby.j.li@nd.edu](toby.j.li@nd.edu) (T. J. Li); [LC@raylc.org](LC@raylc.org) (R. LC) © 2022 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4 .0 International (CC BY 4 .0) . CEURWorkshopProceedings http://ceurISSN 1613-ws-0073.org CEUR Workshop Proceedings ([CEUR-WS.org](CEUR-WS.org)) ing how users perceive the AI used for text generation and how users interact with AI in the creative writing process[18, 19, 20] . Most designs consider collaborative creative writing systems with AI as the user’s assistant, such as supplementing the user’s unfinished sentences or providing users with suggestions for writing [10, 11, 13] . We seek to explore how an AI system may play a more active role in co-

FAQs

What is the main goal of the study on human-AI collaborative fiction writing?

The study investigates how human writers collaborate with ML systems in fiction writing, focusing on interaction capabilities and how these capabilities shape the co-creation experience.

How does the prototype collaborative writing tool work?

The tool uses a web interface where humans and the AI take turns writing to continue the story, offering two modes: Edit Mode and No Edit Mode. Writers can regenerate or choose among AI-generated continuations, and the AI incorporates user changes for the next generation.

What did the preliminary study with 9 users reveal about user perceptions of the AI?

Users drew inspiration from unexpected machine-generated text, expected lower fluency and coherence in machine text when editing was allowed, and perceived the AI as an active writer rather than a passive assistant.

Aug 28, 2026

Uploaded by Evangeline