1 of 11

GenAI in Professional Editing: Goals, Perceptions, and Pedagogy

Brandy Dieterle

Purdue Northwest

brandydieterle.com => writing & research tab has links to slides/script

2 of 11

Introduction

A bit about me:

  • Millennial who grew up with technology,
  • A cautious integrationist.

A refresher on the field:

  • Writing and rhetoric has long seen concern over new technologies.

In this presentation, I will be sharing my experience with integrating GenAI into a graduate level professional editing course.

ThePhoto by PhotoAuthor is licensed under CCYYSA.

3 of 11

Charles Moran (1983):

"Despite the problems inherent in the new technology, word-processing is not going to go away. It will become the norm at colleges and universities, as it is now the norm for professional writers. It will, and should, become part of the writing classroom."

4 of 11

GenAI and Writing

Writing Process: Graham (2023) suggested teaching AI as part of the dialogic and recursive writing process. He noted that AI adds multiple dimensions of recursion involving prompt- engineering, output curation, fact-checking, and revision.

Critical Examination: Byrd (2023) recommended tasks like checking for hallucinations, bias, bigotry, and examining language and textual representations. This helps develop writing skills through critical examination of AI outputs.

Personal Experience: This assignment was my first attempt to critically look at LLM outputs. I found student reflections revealing, noting AI-generated essays were void of meaning, and some encountering 'hallucinations' or made-up information

5 of 11

Project Overview

  • The assignment, 'Editing AI-Generated Content', took place in a 16-week graduate-level course called 'Editing Professional Writing' in Spring 2024.
  • Students used an AI tool to generate an essay, edited it for readability, fact-checked for accuracy and plagiarism, ensured clarity and coherence, and adhered to a style guide.
  • Discussions before and after the assignment captured student perceptions and reflections.

6 of 11

Findings

  • Surface-Level Competence vs. Lack of 'Human' Element
    • AI produced grammatically correct but lifeless text. Students noted the lack of voice, emotion, originality, and nuanced thought. For example, Student 1 described AI as 'kind of cold'.
  • Challenges with Accuracy and Fact- Checking
    • Issues with factual accuracy and 'hallucinations' were frequent. One student found ChatGPT incorrectly agreeing with her misconception about plankton.
  • Difficulties with Style, Tone, and Nuance
    • AI struggled with tone, style guides, and producing engaging prose. For instance, ChatGPT’s tone was described as 'dry' and 'jagged and unclean'.
  • Inaccuracies with Word Count and Formatting
    • AI had problems with word count and formatting, such as providing significantly shorter essays than requested and disjointed paragraphs.

7 of 11

Changing Role of Editors

  • Adapting to AI-generated Content:
    • Editors found themselves focusing more on structure, clarity, and argumentation, moving away from primarily grammar and mechanics.
  • Extensive Rewriting:
    • Some students felt more like writers due to the need for significant rewriting and expansion of AI-generated content.
  • Ethical Considerations:
    • Concerns about plagiarism and AI's lack of automatic citations were noted, necessitating explicit prompting for citations.
  • Emotional Aspect:
    • Some students found editing AI less emotionally charged than peer work, feeling a 'weight lifted' when critiquing non-human text.

8 of 11

Pedagogical Implications

  • Suggests a shift towards macro-level editing skills like structure, clarity, and argumentation.
  • Necessitates instruction in fact-checking, bias detection, and source verification.

The evolving role of editors includes active refinement of AI-generated content. Prompting is crucial for quality output. Integrating GenAI provides opportunities for practicing editing skills without the emotional charge of peer critique.

9 of 11

AI in Editing: Potential and Limitations

  1. Potential Uses:
    • Some students saw AI's potential for brainstorming, structuring, and generating initial drafts. However, its limitations in creative, nuanced, and factually accurate writing were emphasized.
  2. Editing Skills:
    • Editing AI requires unique skills in critical evaluation, fact-checking, rewriting, and macrolevel focus.
  3. Prompting Quality:
    • The quality and specificity of prompts significantly impact AI output. Effective prompting is a key skill.
  4. Ethical Use:
    •  Teaching about proper attribution, fact-checking sources, and ethical use of AI in writing is crucial

10 of 11

Conclusion

  • Integrating GenAI in writing pedagogy highlights the importance of macro-level editing skills and addressing ethical considerations.
  • Students' experiences highlight AI's current limitations and potential in writing and editing.
  • The evolving landscape necessitates adapting pedagogical approaches.
  • My reflections, supported by Notebook LM, emphasize both the opportunities and challenges AI brings to writing education.
  • This talk itself leveraged AI, raising questions about intellectual contribution and the balance between human and AI inputs.

11 of 11

References

​

  • Byrd, A. (2023). Where we are: AI and writing truth-telling: Critical inquiries on LLMs and the corpus texts that train them. In Composition Studies (Vol. 51).
  • Graham, S. S. (2023). Post-process but not post-writing: Large Language Models and a future for composition pedagogy. Composition Studies, 51(1), 162–168.
  • Moran, C. (1983). Word processing and the teaching of writing. English Journal, 72, 113-115.

Personal Note On AI Usage:

  • I used Notebook LM to summarize prompts for the main assignment and two discussion boards and to generate talking points in the findings and pedagogical implications based on my own notes and reflections.

​