Towards Full Authorship with AI: Supporting Revision with AI-Generated Views

TL;DR

The paper addresses preserving authorial autonomy in AI-assisted writing by shifting from system-originated text to user-originated revision. It introduces Textfocals, a Microsoft Word add-in that presents LLM-generated views—summaries, questions, and advice—via a contextual sidebar and provides prompt scaffolding to observe rather than generate text. The study demonstrates that these LLM views can help writers develop underdeveloped ideas, tailor writing to an audience, and clarify prose, while also identifying design challenges in navigation, scoping, prompt engineering, and context management. By enabling reflection and discovery within the writer's own draft, Textfocals contributes to a design space for AI-powered writing tools that maintain authorship integrity and support independent revision.

Abstract

Large language models (LLMs) are shaping a new user interface (UI) paradigm in writing tools by enabling users to generate text through prompts. This paradigm shifts some creative control from the user to the system, thereby diminishing the user's authorship and autonomy in the writing process. To restore autonomy, we introduce Textfocals, a UI prototype designed to investigate a human-centered approach that emphasizes the user's role in writing. Textfocals supports the writing process by providing LLM-generated summaries, questions, and advice (i.e., LLM views) in a sidebar of a text editor, encouraging reflection and self-driven revision in writing without direct text generation. Textfocals' UI affordances, including contextually adaptive views and scaffolding for prompt selection and customization, offer a novel way to interact with LLMs where users maintain full authorship of their writing. A formative user study with Textfocals showed promising evidence that this approach might help users develop underdeveloped ideas, cater to the rhetorical audience, and clarify their writing. However, the study also showed interaction design challenges related to document navigation and scoping, prompt engineering, and context management. Our work highlights the breadth of the design space of writing support interfaces powered by generative AI that maintain authorship integrity.

Figures (2)

• Figure 1: Our Textfocals prototype implemented as a Microsoft Word add-in. Users can interact with the prototype as a "Taskpane" in their Microsoft Word document, as shown in (A). Textfocals includes grouped buttons of predefined prompts, a prompt editor, and cards (views), as shown in (B). The chatbot prototype developed for the pilot user study has a conventional chatbot UI, as shown in (C).
• Figure 2: An overview of a user’s interaction with Textfocals: the user selects or customizes a predefined prompt. This prompts the LLM to generate views that the user can utilize for reflection and discovery in their writing process. Ultimately, this may result in the user’s revision of the composition.