Writing with Artificial Intelligence

Writing with Artificial Intelligence is an undergraduate course that focuses on the question of human agency. Students learn critical AI literacies, which the MLA-CCCC Joint Task Force on Writing and AI has identified as foundational to contemporary literacy. Using these critical frameworks, they identify and debunk “AI hype” and methodological errors in media articles and academic research on AI. They analyze ethical issues associated with large language models (LLMs), including plagiarism, academic integrity, modern copyright, open copyright, privacy concerns, and deepfakes. Students write with generative artificial intelligence tools to compose memos, summaries, articles, songs, short stories, poems, deepfakes, misinformation campaigns, images, videos, and bots. Thereafter, they conduct qualitative, empirical research to investigate how those tools impinged on their learning, thinking, self-expression, and creativity. Finally, students speculate about how the rise of "superintelligence" (Aschenbrenner 2024) and artificial general intelligence will challenge humans to reimagine what it means to be human.

Writing with Artificial Intelligence - WritingCommons

Course Description1

The rapid progression of generative artificial intelligence raises profound questions about the future of human agency, creativity, authorship, copyright, intellectual property, and the role of writing in learning and knowledge production. In 2022 Anderson and Rainie, in partnership with the Pew Research Center, surveyed 540 “technology innovators, developers, business and policy leaders, researchers, academics and activists.” They found that 56% of the experts believe AI will limit human agency, expression, and creativity. 

In a subsequent study, Rainie and Anderson surveyed and canvassed, in partnership with Elon University’s Imagining the Digital Future Center,” hundreds of global technology experts to investigate their thoughts on AI and the future of humanity. One of the five major concerns expressed by these global experts was the fear that “humanity could be greatly enfeebled by AI:”

A share of these experts focused on the ways people’s uses of AI could diminish human agency and skills. Some warned it will nearly eliminate critical thinking, reading and decision-making abilities and healthy, in-person connectedness, and lead to more mental health problems. Some said they fear the impact of mass unemployment on people’s psyches and behaviors due to a loss of identity, structure and purpose. Some warned these factors combined with a deepening of inequities may prompt violence (Rainie and Anderson 2024).

More recently, Leopold Aschenbrenner, a researcher at OpenAI, has speculated, based on historical trends, that AI may reach GAI “superintelligence” by 2027 (see Situational Awareness). At that point Aschenbrenner predicts generative AI tools will be able to write as well as Ph.Ds in computer science — or any other topic in any language. AI systems will no longer need humans to engage in inquiry. Rather, given their thirst for knowledge, they will be defining new problems, creating new medicines, surgical procedures, and ways of knowing. They will be the smartest being on the planet.

This is a project by Aschenbrenner about when GAI systems will attain superintelligence.
Once AI systems attain superintelligence will humans become lazy expecting AI to do all of the writing and thinking Ultimately if AI takes puts a lot of people out of work what will those folks do <a href=httpssituational awarenessaiwp contentuploads202406situationalawarenesspdf target= blank rel=noopener title=>Aschenbrenner L 2024 June p 48<a>

This course tackles this question of agency — of whether the American people can be “masters of their technology and not its unthinking servants” (NEH).  In response to eight creative challenges, students write with and about AI technologies. Experimenting with a variety of AI tools, they create arguments, songs, images, videos, deepfakes. Subsequently they reflect on their works and processes, questioning how the AI tools constrained and enhanced their self expression, thinking, and creativity:

  1. Key Benefits of Writing Without AI for Students
  2. Create a Prompt Engineering Infographic Tutorial
  3. Build a Custom Chatbot
  4. Practice Critical AI Literacies
  5. The Future of Writing: Postplagiarism & Hybrid Writing?
  6. Research How AI Constrains & Enhances Your Agency as a Human Being
  7. Research Deepfakes & Misinformation
  8. Imagine the Digital Future (Optional Project)

Students will complete quizzes and assignments in class that must be written in class. They will also be assigned readings and required to annotated those readings in Perusall, a social annotation app that the university makes freely available to students inside Canvas.

Altogether, the readings, in class writing workshops and discussions, and writing assignments challenge students to

  1. review critical AI literacies — critical intellectual frameworks — that help writers identify “AI hype” in published media articles on AI
  2. analyze ethical issues associated with large language models (LLMs), including plagiarism, academic integrity, modern copyright, open copyright, privacy concerns, and deepfakes
  3. experiment with generative artificial intelligence (GAI) tools — such as ChatGPT-4o, Claude 3 Opus, Gemini 1.5, Microsoft Copilot, Perplexity, Midjourney, Deepfake Web, Suno, and PoeAI — to compose annotations, summaries, articles, songs, short stories, poems, deep fakes, images, videos, and bots
  4. conduct qualitative, empirical research to investigate how the use of GAI tools affects their learning processes, thought patterns, self-expression, and creativity
  5. write brief articles and memos in response to creative challenges and course readings.

Regarding scope, this course will not address the environmental consequences associated with AI usage — although that, most assuredly, an important concern. Additionally, we do not address privacy and surveillance issues, given time limitations

Creative Challenges – Writing Assignments

Learning Objectives – Course Goals

  1. Demonstrate the ability to find and use AI tools for specific purposes, genres, and media while maintaining authorial agency and voice.
  2. Learn to use GAI tools critically to accomplish aims
  3. Learn to employ critical frameworks to identify and debunk AI “ hype”
  4. Learn strategies for preserving agency, human creativity, critical thinking, and decision-making while using AI to create, research, and learn
  5. Understand the ethical issues associated with large language models 
  6. Demonstrate the ability to critically evaluate AI-generated texts and identify when human intervention and original thought are necessary
  7. Demonstrate the ability to develop strategic AI prompts to locate and evaluate information, develop ideas, structure arguments, and refine drafts, all the while maintaining authorial control

Required Texts

  1. All texts will be provided through Canvas/Perusall or will be freely available on the internet.
  2. GCF Global. Google Drive and Docs
    This is a thorough, free guide to using gDocs. You’ll need this resource if you are unsure how to create and share gDocs.
  3. Writing Commons
    You can block the ads by adding Adblock Plus, a free Chrome extension. The ad blocker works great.
Just as Orville Wright’s first flight marked the dawn of a new era in aviation, today’s AI, still in its infancy, is rapidly evolving and poised to transform our future. Soon — maybe as early as 2027 — AI systems will achieve superintelligence: They will surpass Ph.D. researchers in intelligence and conduct their own independent research. Humans will no longer be the smartest beings on the planet (Aschenbrenner 2024).

Grading Contract

Students’ grades will be based on their labor over the semester. This approach is called “labor-based contract grading.” Ideally, contract grading frees you up to try new things because you won’t be penalized for taking risks. In fact, I strongly (!!) encourage you to try new things and push yourself. Growth and strength result from struggle and working through confusion.

Contract Grading – UnGrading Resources

  1. Contract Grading – So Your Instructor Is Using Contract Grading
  2. Labor-Based Grading Resources by Asao Inoue
  3. Labor-Based Grading Contracts: Building Equity and Inclusion in the Compassionate Writing Classroom, 2nd Edition

Canvas Workaround

Assignments in Canvas will be marked as “Complete” or “Incomplete.” Nonetheless, Canvas will show you a percentage in your “Grades” view. However, because we are using a labor-based contract the Canvas grade percentage means nothing. Absolutely nothing. Ignore it. Any cumulative percentage that Canvas might show you is meaningless.

Grading Criteria

  1. You earn a score of completion on an assignment by completing it as described in the assignment description. Your submissions should demonstrate you have read the readings associated with an assignment. 
  2. You earn an incomplete by failing to submit an assignment, by submitting an assignment that cannot be opened/read or, when required, commented on; by submitting an assignment that is not responsive to the assignment prompt; or by submitting work that is writer-based as opposed to reader-based — i.e., writing that is sloppy, writing that is so personalized, so idiosyncratic, that readers cannot successfully interpret it. Discourse may be called writer-based when it lacks an organizational structure other than a stream of consciousness, when it departs so significantly from standard written English that readers cannot decipher what the writer is saying. Furthermore, writing that does not in my opinion seem like prose from one human to another human will be marked as incomplete.

To earn an A in this course, you will need to

  1. receive a complete on all eight creative challenges
  2. complete all of the annotations/readings assigned in Perusall
  3. attend all but three classes

To earn a B in this course, you need to

  1. receive a complete on the first seven creative challenges
  2. not be noticeably late to in-person class meetings more than twice.
    • Three late arrivals will constitute one missed assignment. 
    • Five late arrivals will constitute a second missed assignment
    • Following five late arrivals, each lateness will result in a course grade deduction. So, for instance, a final grade of a B will become a C.

To earn a C in this course, you need to

  1. Meet the expectations for a B, yet receive a complete on six of the first seven creative challenges

To earn a D in this course, you need to

  1. Meet the expectations for a B, yet receive a complete on six of the first seven creative challenges.

In summary, your grades are based primarily on your labor as opposed to the quality of your work.

Grading FAQs

If you are grading based primarily on labor rather than quality and assigning “complete” or “incomplete” grades, what sort of critical feedback can I expect to receive?

Conventions

Depending on the rhetorical context, I’ll consider the conventions that govern academic or professional writing:

  1. Academic Writing – How to Write for the Academic Community
  2. Professional Writing – How to Write for the Professional World

Audience Awareness

I will assess whether your work is responsive to the needs and interests of its target audience (e.g., readers, listeners, or users). As NCTE’s (National Council of Teachers of English) Position Statement on “Understanding and Teaching Writing: Guiding Principles” (Adler-Kassner et. al. 2018) points out, audience awareness is a critical concern of writers during composing (along with purpose and context):

When writers produce writing, they take into consideration purposes, audiences, and contexts. This leads them to make intentional choices about the elements that go into writing:

  1. content (the subject or focus of the writing);
  2. form (the shape of the writing, including its organization, structure, flow, and composition elements like words, symbols, images, etc.);
  3. style and register (the choice of discourse (aka writing style] and syntax used for the writing, chosen from among the vast array of language systems [often called “dialects”] that are available for the writer); and mechanics (punctuation, citational style, etc.)” (“Understanding” 2022).

Style

I will assess whether the writer(s) has adopted an an appropriate writing style given the rhetorical situation. Are the writer’s appeals to ethos and pathos appropriate given the audience? Have they established a consistent voice, tone, and persona? To assess whether the text is writer-based or reader-based based, I will evaluate its clarity, brevity, coherence, flow, inclusivity, simplicity, and unity

Content & Critical Thinking

I will evaluate whether the writer has provided the evidence and reasoning readers need to correctly interpret the work. Regarding evidence, is the content responsive to what the audience knows/feels about the topic? Has the writer created an authoritative text by providing  a consistent credible voice, tone, and persona? Have they employed the information literacy conventions academic and professional readers expect? For instance, have they provided the sources and details readers need to assess the credibility of their claims? Additionally, I will assess whether the writer has maintained a consistent line of inquiry or analysis throughout the paper. Has the writer demonstrated a clear progression of ideas, where each new piece of information logically builds on the previous one.  This approach ensures that the reasoning is clear and coherent, effectively addressing the thesis or research question and demonstrating thorough content and critical thinking.

Organization

An illogical progression or lack of cohesiveness will hinder clarity and undermine the effectiveness of the writing. A well-organized paper demonstrates an understanding of the rhetorical situation and audience’s needs, resulting in a clear and compelling piece. Hence, I will assess check the document for logical flow — whether the writer maintains a consistent line of inquiry or analysis throughout the paper. This involves ensuring that every section and paragraph supports the central focus—the thesis, hypothesis, or research question that drives the narrative or argument. In other words, I’ll consider whether the writer has structured their work with a clear and logical progression of supporting points, ensuring cohesiveness and unity throughout. This involves ensuring that every section and paragraph supports the central focus — the thesis, hypothesis, research question that drives the narrative.This also involves ensuring each new idea builds logically on the previous one, adhering to the given-to-new contract. I’ll also consider whether deductive and inductive reasoning are applied appropriately, depending on the nature of the argument or narrative. Headers should be used effectively to make the content scannable.

Design

I will assess whether the writer has effectively applied key design principles — proximity, alignment, repetition, and contrast — to enhance the clarity and impact of their work. I will question whether the text demonstrates an understanding of visual rhetoric and the power of visual language. This includes using images, graphs, and other visualizations to support and enhance the written content, making complex information more accessible and engaging. Headers, bullet points, and other formatting tools should be used effectively to make the document scannable and user-friendly. 

Acknowledgements1

For helping me develop the course readings and assignments, I thank Ilene Frank (HCC) librarian and friend extraordinaire;  Professors Whitney Gregg-Harrison (University of Rochester); Abram Anders (Iowa State University); and Anna Mills (Cañada College), and Troy Hicks (Central Michigan University). Thank you, colleagues, for being so inspirational and sharing your expertise. For help with the contract grading I’m using in this course, I thank Heather Shearer (Teaching Professor at UC Santa Cruz). Now that GAI tools are so widely available and given it’s impossible to police their use, “ungrading” is, in my opinion, the only sensible way forward.