AI Text Generators and Teaching Writing: Starting Points for Inquiry

 

Introduction

What do teachers who assign writing need to know about AI text generators? How should we change our pedagogical practices, given the recent advances in AI Large Language Models (LLMs) such as OpenAI's ChatGPT? How should teachers participate in shaping policies around these technologies in our departments, institutions, and society?
As teachers who assign writing, we need to reckon with the possibility that our students may use writing generators to produce essays or parts of essays. These tools do not produce writing copied from human sources; the AI origin appears to be not reliably detectable either by human readers or by software. 
To shape our individual and institutional responses to this new technology, writing teachers and scholars need more information about the kinds and quality of AI-generated text we can expect in response to common types of essay prompts. For example, we may want to design prompts that AI text generators are unskilled at performing, or we may want to find ways to use these generators pedagogically. Either way, as faculty responsible for teaching writing as impactful, ethical intellectual activity, we need to know what AI generators are capable of. And we need to understand the forms of bias and error that emerge in AI-generated writing.

Shared Resources

• Resources on language models, learning, and teaching
• Experimental essays created with OpenAI's GPT-3
• Questions for writing teachers to consider
• Strategies for mitigating harm associated with language models and writing

Opportunities to Engage and Participate

• Comment in the margins and read others’ comments on articles on AI text in education with Hypothesis. Tag your comments ChatGPTedu and AItextedu.
• Comment on the sample experimental essays.
• Join the listserv AIGeneratorsandTeachingWriting. 
• Share your own found or created examples of AI-generated essays.
• Suggest additional resources, questions, or strategies for teaching policies and practices. 
• Add to the list of strategies to mitigate harm involved with AI-generated writing.

The resources collected here serve as catalysts for inquiry, discussion and collaborative research as we respond to this major change in the kind of writing assistance available to our students.

Suggested Citation

Mills, Anna (Curator). AI Text Generators and Teaching Writing: Starting Points For Inquiry. 2022. Last Updated February 14, 2023.

Authorship and Credits

Anna Mills (armills@marin.edu) teaches English at College of Marin and previously taught at City College of San Francisco for 17 years. She is the author of an Open Educational Resource (OER) textbook, How Arguments Work: A Guide to Writing and Analyzing Texts in College, which has been praised on OER review sites and used at over 45 colleges. She is the recipient of an Open Education Research Fellowship and currently serves as the English Discipline Lead for the Academic Senate of the California Community Colleges OER Initiative.  Anna earned a master’s degree from Bennington College in Writing and Literature with a focus on nonfiction writing. Her essays have appeared in journals such as The Writer's Chronicle, The Sun, and Salmagundi. Recently, she has focused on exploring how writing instructors can respond to the accessibility of large language models. Follow her at @EnglishOER.