CMSC 33231

Topics in HCI: Writing with AI

Spring 2026  •  University of Chicago
Writing with AI — a human hand and a robotic hand, both writing with a pen
Instructor Mina Lee
Time Friday 1:30-4:30 PM
Location DSI 1035460 S University Ave.
Please enter through the South and North entrances of the building.

Overview

How is AI changing what we write, how we write, and who we are as writers?

This interdisciplinary seminar course explores the topic of AI and writing at the intersection of human-computer interaction, natural language processing, computational social science, education, and beyond. Students will review prior research, identify current opportunities and risks for AI writing assistants, and envision future designs in the area. The course includes weekly paper presentation and discussion as well as hands-on activity to experience and reflect on writing with AI.

  • Prerequisite: None, though basic familiarity with AI is recommended (e.g., enough to follow news articles and understand common terms, such as LLMs, APIs, and UI/UX), as well as comfort trying out different AI systems including those that require a terminal (command line interface), such as Claude Code. Open to students from any discipline.
  • Elective: Counts toward the CS elective requirement (AI/ML and HCI).
  • Auditing: Auditors are welcome with instructor permission. Please email to request and come to the first class.
  • Enrollment: No enrollment is allowed after Week 2.
⚠ This syllabus is tentative and subject to change. This is the first time this course is being offered, and we will refine it together as we go. Your feedback and suggestions are always welcome throughout the quarter.

Format

Each class meeting will roughly follow this format (Week 2-8):

Contents Duration
Part 0 Pulse: Writing with AI in the news 1:30-1:40 PM (10 min)
Part 1 Deep dive: Writing with AI in research 1:40-2:50 PM (70 min)
Paper presentation 1 20 min
Paper presentation 2 20 min
Discussion 30 min
Part 2 Practice lab: Writing with AI for your own use 3:00-4:00 PM (60 min)
Instructions & Demonstration 20 min
Writing 20 min
Reflections 20 min
Part 3 Open studio (Flexible time for feedback, check-ins, and continued work) 4:00-4:20 PM (20 min)

Course Schedule

The topics are organized based on the five aspects of AI writing assistants in Lee et al. (CHI 2024):

  • Task: Writing stages, contexts, and purposes that writing assistants aim to support
  • User: Characteristics and preferences of users of writing assistants
  • Technology: Building blocks of developing underlying models that power writing assistants
  • Interaction: Diverse interaction paradigms and essential user interface components of writing assistants
  • Ecosystem: Broader context in which writing assistants operate, including economic, social, and regulatory considerations that impact how these systems are deployed, used, and evolved

Readings are subject to change and will be finalized two weeks before each class meeting. Finalized readings are marked with ✓ next to the week number.

Show:
Deep Dive Practice Lab
Week 1✓
Mar 27		 Background: What is writing with AI?

Papers

  1. CoAuthor: Designing a Human-AI Collaborative Writing Dataset for Exploring Language Model Capabilities (Lee et al., CHI 2022)
  2. A Design Space for Intelligent and Interactive Writing Assistants (Lee et al., CHI 2024)

Additional references

  1. Ghosts by Vauhini Vara [Link]
  2. Stanford HAI Seminar with Vauhini Vara: The Impact of AI on Writing [Video]
  • Identify a writing task and a concrete project you want to try writing with AI this quarter.
  • Reflect: Why are you interested in using AI for this task? What does that tell you about the task and your relationship to it?
  • Capture your current approach: What AI systems do you use? How often, for what tasks, and in what ways?
Week 2✓
Apr 3		 User: What impact does writing with AI have on our writing and judgment?

Papers

  1. Biased AI Writing Assistants Shift Users' Attitudes on Societal Issues (Williams-Ceci et al., Science Advances 2026)
  2. Writing with AI Can Reduce Gender Bias in Hiring Evaluations (Liu et al., CHI 2026)

Additional references

  1. Does Writing with Language Models Reduce Content Diversity? (Padmakumar and He, ICLR 2024)
  2. AI Suggestions Homogenize Writing Toward Western Styles and Diminish Cultural Nuances (Agarwal et al., CHI 2025)
  3. How LLMs Distort Our Written Language (Abdulhai et al., 2026)
  • Experiment with prompting strategies from the readings and compare outputs. Does the AI steer your writing in a particular direction (e.g., stylistically or ideologically)?
  • Reflect: How do the suggestions make you feel about your own writing?
  • Do you want to steer the AI toward your voice and values, or do you want to use the AI's suggestions to challenge and expand your thinking? Apply prompting techniques to your writing task to achieve your desired outcome.
Week 3✓
Apr 10		 Task: How should AI support different kinds and processes of writing?

Background

  1. A Cognitive Process Theory of Writing (Flower & Hayes, College Composition and Communication 1981)

Papers

  1. A Design Space for Writing Support Tools Using a Cognitive Process Model of Writing (Gero et al., In2Writing 2022)
  2. Co-Writing with AI, on Human Terms: Aligning Research with User Demands Across the Writing Process (Reza et al., CSCW 2025)

Additional references

  1. The Metacognitive Demands and Opportunities of Generative AI (Tankelevitch et al., CHI 2024)
  • Explore various AI writing assistants designed for specific writing processes (e.g., ideation, drafting, or reviewing).
Week 4✓
Apr 17		 User: What impact does writing with AI have on our brains?

Papers

  1. The AI Memory Gap: Users Misremember What They Created With AI or Without (Zindulka et al., CHI 2026)
  2. Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task (Kosmyna et al., 2025) *Abstract, Summary of Results, Experimental Design (22-51p), Discussion (143-151p)

Additional references

  1. A.I. Is Homogenizing Our Thoughts (The New Yorker, 2025) [Article]
  2. The Memory Paradox: Why Our Brains Need Knowledge in an Age of AI (Oakley et al., 2025)
  3. Thinking—Fast, Slow, and Artificial: How AI is Reshaping Human Reasoning and the Rise of Cognitive Surrender (Shaw and Nave, 2026)
  • Write with the AI for 10 minutes, then write without AI for 10 minutes. Compare the two processes and outputs.
Week 5
Apr 24		 Interaction: What new ways of writing does AI make possible?

Papers

  1. Collage is the New Writing: Exploring the Fragmentation of Text and User Interfaces in AI Tools (Buschek, DIS 2024)
  2. Textoshop: Interactions Inspired by Drawing Software to Facilitate Text Editing (Masson et al., CHI 2025) (Masson et al., CHI 2025)

Additional references

  1. Script&Shift: A Layered Interface Paradigm for Integrating Content Development and Rhetorical Strategy with LLM Writing Assistants (Siddiqui et al., CHI 2025)
  • Explore multiple AI writing assistants with novel interaction paradigms (e.g., direct manipulation, multi-modal).
Week 6
May 1		 Technology: How good can AI write, really?

Papers

  1. Readers Prefer Outputs of AI Trained on Copyrighted Books over Expert Human Writers (Chakrabarty et al., 2025) *Main text (1-13p)
  2. Collaborative Document Editing with Multiple Users and AI Agents (Lehmann et al., CHI 2026)

Additional references

  1. Who's a Better Writer: A.I. or Humans? Take Our Quiz. by The New York Times [Article]
  2. Thinking with AI: Machine Learning the Humanities (Bajohr, 2025) [Book] *Chapter 3
  3. On the Opportunities and Risks of Foundation Models (Bommasani et al., 2021) *Section 1
  • Apply advanced prompting techniques and agentic workflow to your writing task.
Week 7
May 8		 Technology: What are ethical dilemmas around data and models that power AI writing assistants?

Papers

  1. On the Dangers of Stochastic Parrots: Can Language Models Be Too Big? (Bender et al., FAccT 2021)
  2. Creative Writers' Attitudes on Writing as Training Data for Large Language Models (Gero et al., CHI 2025)

Additional references

  1. The Foundation Model Transparency Index (Bommasani et al., TMLR 2024)
  2. More Than Words: How to Think About Writing in the Age of AI by John Warner [Book] *Chapter 2
  • Find free and open-source data and build a custom AI writing assistant through fine-tuning or in-context learning.
Week 8
May 15		 Ecosystem: When should we disclose AI use in writing and how?

Papers

  1. Which Contributions Deserve Credit? Perceptions of Attribution in Human-AI Co-Creation (He et al., CHI 2025)
  2. What Influences Readers' and Writers' Perceived Necessity of AI Disclosure? (Fang et al., FAccT 2026)

Additional references

  1. A.I. Is Writing Fiction. Publishers Are Unprepared. by The New York TimesTuhin Chakrabarty, a professor of computer science at Stony Brook University, used Pangram to check more than 14,000 self-published novels on Amazon for A.I. writing. The program found that nearly 20 percent of the novels had been substantially written by A.I. Looking mostly at novels released between 2024 and 2025, Chakrabarty saw a 41 percent jump year-over-year in how many novels in his random sample contained a large amount of A.I. generated text, he said. [Article]
  2. AI Attribution Toolkit by IBMThis is what I used to generate the attribution mark for the image at the top of this page! [Tool]
  3. How AI disclosures in news help — and also hurt — trust with audiences by Trusting News [Article]
  • Explore various stylometry metrics and learned models to analyze the degree of AI involvement in your writing and how that correlates with your own perception of AI contribution.
  • Experiment with different AI disclosure practices in your writing (e.g., AI Attribution Toolkit, Process Feedback) and design your own approach.
Week 9
May 22		 Ecosystem: What is the future of writing and writing education?

Papers

  1. Future of Writing & Writing Instruction (Kaufer and Ishizaki, 2025)
  2. From Crafting Text to Crafting Thought: Grounding AI Writing Support to Writing Center Pedagogy (Liu et al., CHI 2026)

Discussion

  1. How necessary are writing skills in the future?
  2. What is gained and lost through the use of AI in writing?
  3. What should writing education prioritize in the age of AI?
  • (Student presentation)
Week 10
May 29		 No class

Coursework

1. Weekly Reading and Participation

Writing with AI in the news (optional)

If you come across an interesting news story or development related to writing with AI, share it in the #news channel on Discord and bring it up during Part 0 of class. It doesn't have to be directly tied to the week's topic, but you're always welcome to connect it to the topic (optional).

Paper reading

Each week, we read 2 papers together.

Before class: Read all assigned papers and engage on the social reading platform (Hypothesis)

  1. Leave at least one comment on each paper (due Wed at noon)
    ▶ Examples of what to comment on
    • Something that was thought-provoking from the paper and why.
    • Question(s) you have about the paper.
    • An argument you would like to challenge about the paper.
    • A weakness or limitation of the paper.
    • A strength of the paper that you would want to explore further or incorporate into a practice lab.
    • A part of the paper where you could extend future work.
    • The applicability of an argument or idea from the paper that is related to recent events.
    • Something new you learned from the paper that you would incorporate into your personal AI-assisted writing practices.
  2. Reply to at least one classmate's comment on each paper (due Fri at noon)
  3. Read additional references (optional)

During class: Participate fully in discussions and activities. Come prepared to share your perspectives on the readings.

Practice lab

Each week, we do a practice lab together during class. You will begin working on it during class and finish it on your own afterward.

During class: Start the practice lab activity guided by the presenting team.

After class: Complete and submit your practice lab on Canvas (due next Wed at noon)

2. Team Presentation

Once a quarter, your team (2-3 students) leads either the Deep Dive or the Practice Lab portion of a class session.

Option A: Deep Dive

The team presents and facilitates discussion of the week's assigned papers. If the team wants to suggest alternative papers, please reach out to the instructor at least two weeks in advance.

One week prior to your presentation:

  • Submit your draft materials (slides and discussion questions) to Canvas (due Fri at noon the week before)
  • Meet with the instructor to review (during Part 3: Open studio of the class the week before)

On the day of your presentation:

  • Submit your final materials to Canvas (due Fri at noon)

Tip: During the week of your presentation, keep an eye on the paper reading comments as they come in. Try to address questions classmates raise and incorporate their requests into your presentation and discussion planning.

Paper presentation (20 min per paper)

  • Cover the paper's motivation, goal, approach, findings, and key takeaways
  • Critically evaluate the paper (e.g., assess strengths and limitations, appropriateness of methods, validity of claims)
  • Connect the paper to current events, ongoing debates, or your own experience
  • Take questions

Paper discussion (30 min)

  • Briefly highlight key themes, tensions, or open questions raised by the papers
  • Prepare 2-3 discussion questions to facilitate thoughtful critique and reflective engagement
  • Foster inclusive participation; guide the conversation while staying open to unexpected directions
  • Summarize the discussion at the end

One week after your presentation:

  • Submit a report summarizing the presentations, social reading comments, key points from the discussion, reflections, and any additional references mentioned during class (due Wed at noon the week after)

Tip: I recommend having one team member take notes during the discussion while the other leads it.

Option B: Practice Lab

The team experiments with 10 different ways to write with AI related to the week's topic (see the "Practice Lab" tab in Course Schedule for a rough idea), selects the 2-3 most effective or thought-provoking ones, and designs a practice lab for the class to experience them. The team can define "ways to write with AI" broadly to include different tools, prompts, workflows, etc.

The practice lab should be designed to provide students hands-on experience with writing with AI, encourage reflection on how AI can be used effectively and responsibly, and help everyone optimize a writing task of their choice.

One week prior to your presentation:

  • Share your draft materials (summary of experiments and practice lab materials) to Canvas (due Fri at noon the week before)
  • Meet with the instructor to review (during Part 3: Open studio of the class the week before)

On the day of your presentation:

  • Submit your final materials to Canvas (due Fri at noon)
  • Set up any necessary materials for the practice lab in Google Drive (due Fri at noon)

Tip: During the week of your presentation, keep an eye on the paper reading comments as they come in. Try to address questions classmates raise and incorporate their requests into your practice lab design and activities.

Instructions & Demonstration (20 min)

  • Introduce 2-3 ways to write with AI that you found effective in your experiments
  • Demonstrate them with live examples
  • Provide guidance and tips for using AI effectively and responsibly

Writing (20 min)

  • Students apply the ways to their own writing using AI
  • Take questions and support students as needed

Reflections (20 min)

  • Students share their work and observations
  • The team synthesizes key takeaways and concludes with a brief closing thought
  • The team shares additional resources for further exploration

One week after your presentation:

  • Submit a report summarizing the lab activities, key observations from the writing session, reflections, and any additional references mentioned during class (due Wed at noon the week after)

Tip: I recommend having one team member take notes during the lab while the other leads it.

3. Class Project (elective credit only)

Choose one project from the options below. Projects are individual (no team component) and are designed to accommodate diverse backgrounds and interests.

Tip: Aim for ideas that are thoughtful, critical, and generative, not just polished outputs. If you are interested in expanding your project into a workshop or a full research paper, discuss your ideas with the instructor.

Project A: Write a Tutorial

Create a public-facing tutorial that demonstrates how to optimize a specific writing task using AI writing assistants. Your tutorial should be clear, actionable, and grounded in hands-on experience. It should address a task of moderate to high complexity and go beyond basic prompting by incorporating agentic or multi-step workflows.

Example topics:

  • A 7,500-17,500-word creative story pipeline that survives multiple rounds of planning, drafting, and existential revision with AI
  • A reflective diary workflow that experiments with new ways of thinking and writing, potentially using multiple writing assistants and analysis of past entries
  • An intelligent email assistant that accesses prior emails, adapts to your writing style and recipients, and diplomatically says “per my last email” without starting a war
Project B: Research Proposal

Write a research proposal that explores a specific question at the intersection of AI and writing. Your proposal should clearly define the problem, connect it to existing research, and outline a feasible approach to investigating it. Include the following sections: Introduction, Background & Related Work, Research Questions, Proposed Methods, and Expected Outcomes.

Example topics:

  • The impact of AI writing assistants on writing skill development
  • AI writing assistance for specific populations (e.g., non-native speakers, people with dyslexia)
  • Ethical considerations in AI-assisted writing (e.g., copyright, attribution)
  • AI and the future of writing
Project C: Design an AI Writing Assistant

Design and prototype a new AI writing assistant or workflow that explores new possibilities for how people write with AI. Your project may address an existing need, reimagine current practices, or experiment with bold and unconventional ideas. Ground your design in a clear vision of how the assistant would be used and what it enables.

Project D: Project of Your Choice

Propose and complete a project related to AI and writing that aligns with your interests. This could build on an existing research project, explore a specific domain, or take a creative or experimental approach. You should discuss your idea with the instructor within the first two weeks of class.

Expectations

  • Think beyond the obvious. Be experimental and creative; take intellectual risks. For Project A, the best tutorials reveal workflows that can't easily be Googled; go beyond basic prompting and show something truly exceptional. For Project C, a thin wrapper around an existing API is not enough; aim for something that explores genuinely new possibilities.
  • Make us care. AI can generate similar outputs quickly. In other words, comparable work can often be produced in a single prompt across all project types, and especially for Project B. Explain why your project matters, what makes it interesting, and how it reflects insight that AI alone couldn't produce.
  • Show your thinking process. Document your iterations, decisions, failures, and insights, not just the final output. For Project B, this might mean explaining why you chose this research question over alternatives and what tradeoffs you considered. For Project C, vibe coding is allowed but not sufficient: show what you learned and justify every design choice you made. "Success" doesn't mean a polished outcome, but is a reflection of your critical, iterative, and reflective process.
  • Apply best practices. Reflect on responsible and critical AI use throughout, including AI disclosure (Week 8). For Project C, this also means thinking carefully about the ethical implications of the assistant you're designing.
  • Communicate proactively and clearly. Talk to the instructor and get feedback sooner rather than later, especially if you are working on Project D. In your deliverables, present your ideas, process, and results in a way that is understandable and engaging to a general audience; as always, avoid jargon, or define technical terms in your deliverables when necessary.

Deliverables

  • Week 3: Project proposal (due Fri at noon)
    ▶ Proposal guidelines

    Submit a short proposal (max 1 page, any format) that describes your project idea, explains why it matters, outlines what you plan to do, and reflects on how you'll evaluate its successful completion. Be concise but specific: the goal is to give the instructor enough to assess the topical relevance, complexity and feasibility, and thoughtfulness of your idea. If your project overlaps with other ongoing research or coursework, briefly describe the overlap and how you plan to address it.

    • Project A: Include a rough outline of the tutorial structure and some preliminary examples of the workflow or technique you plan to demonstrate.
    • Project B: Include a rough outline of the proposal sections and a preliminary literature search (e.g., a short list of key papers with a sentence or two on how each relates to your research question).
    • Project C: Include a rough prototype or mockup and a brief description of the design (e.g., what the assistant does, who it's for, and what makes it different and meaningful).
    • Project D: Include a detailed plan (e.g., what you will do week by week, what resources or access you need, and how you will know the project is complete).
  • Week 8: Final deliverables (due Fri at noon)
    ▶ Final deliverable guidelines (tentative)

    Submit your final project in the format appropriate for your project type.

    • Project A: A public-facing document or slides that clearly communicates the tutorial to a general audience.
    • Project B: A research proposal paper (max 6 pages, excluding references) written in Overleaf using the ACM format.
    • Project C: A working prototype or demo, a write-up explaining your design decisions and what you learned, and documentation for the codebase.
    • Project D: Format defined in consultation with the instructor, based on your project plan.
  • Week 9: Poster presentation in class (poster submission due Wed at noon)
    ▶ Poster session guidelines (tentative)

    Each student prints out a poster and presents it during a class poster session (2:40-4 PM) in Week 9. The poster session is divided into two groups: one half presents while the other half walks around, engages with the posters, and votes for their favorites. Groups switch after 40 minutes. Awards are announced at the closing.

    • 2:40-3:20 PM: Session 1: First half presents, second half browses and votes
    • 3:20-4:00 PM: Session 2: Groups switch
    • 4:00-4:20 PM: Closing and awards

    All projects should include a poster. A demo is encouraged where applicable (e.g., Project C). Outstanding projects may be featured on the course website (with your consent).

Grading (tentative)

Seminar (P/F) — 60 pt + up to 2 extra pt
Weekly reading & Participation 30 pt (+ 2 extra pt)
Hypothesis engagement (Week 2-9) (complete/incomplete) 16 pt
Practice lab (Week 1-8) (complete/incomplete) 16 pt
Team presentation 30 pt
Draft materials & review (letter grade) 8 pt
Final materials & presentation (letter grade) 12 pt
Report (letter grade) 10 pt
Elective (A-F) — 100 pt + up to 5 extra pt
Weekly reading & Participation 30 pt (+ 2 extra pt)
Hypothesis engagement (Week 2-9) (complete/incomplete) 16 pt
Practice lab (Week 1-8) (complete/incomplete) 16 pt
Team presentation 30 pt
Draft materials & Review (letter grade) 8 pt
Final materials & Presentation (letter grade) 12 pt
Report (letter grade) 10 pt
Class project 40 pt (+ 3 extra pt)
Proposal (letter grade) 5 pt
Final deliverable (letter grade) 25 pt (+2 extra)
Poster presentation (complete/incomplete) 10 pt (+1 extra)

Resources

There is no required textbook for this course. The following books are related to the topics covered in this course and may be of interest:

AI Policy

AI use is allowed and encouraged. This course is designed to be a safe space to explore AI's capabilities and limitations in writing, so be creative and proactive! Use AI however you find useful, and don't be afraid to push its boundaries. (And yes, I used AI to help draft and revise this very policy.)

That said, you are fully responsible for everything you submit. Your work will be evaluated on its quality, and you are expected to understand and speak to every part of it. If your understanding doesn't match the level of your submission, that's a red flag. Any error, flaw, or unsupported claim is yours to own, regardless of whether AI produced it.

The goal isn't to restrict how you use AI; it's to help you use it responsibly and effectively. Getting impressive output from AI is easy. Knowing when to trust it, when to push back, and how to make it work toward your own learning and fulfillment -- that's the core skill this course is here to cultivate.