Embracing the bot: designing writing assignments in the face of artificial intelligence

Just as pocket calculators, personal computers, and smartphones have posed threats to students learning math skills, artificial intelligence (AI) appears to be the new tool poised to undermine the use of writing assignments to assess student learning.

In November 2022, a tool called ChatGPT made headlines for its ability to “write” any content. As an instructional designer, I immediately heard from anxious faculty that the sky might be falling, and wondered what their chance would be against robots that could write student papers.

After some reflection, I’ve come to believe that, in the long run, worrying about how students will use AI to cheat isn’t the most productive question to focus on. The better question is, even in the age of artificial intelligence, how can we best teach our students? Here are three ways to design writing assignments in the face of the onslaught of AI.

The first way: Ignorance is bliss

In the extreme responses, we have the approach of “ignorance is bliss” and “resistance is useless.” These positions are grouped together because they both prefer to avoid the underlying issue. In the first case, the teacher may simply be unaware that students can now type a writing prompt into a website and copy the answer it generates into a document to send. In the latter case, the teacher may be aware of the AI’s ability to write, but may metaphorically raise his arms to the prevailing notion that he can no longer tell whether or not a student has written a submitted research paper.

At worst, teachers with this mindset can resign themselves to classifying work written by AI and hope that most students will still write their own papers and learn from the notes. For teachers who do assessments to help students develop their writing skills, it would be a waste of time to respond to anything their students haven’t written – and those students won’t invest as much in reviewing comments.

For teachers who recognize the AI’s ability to write a research paper but don’t feel ready to deal with a bot face-to-face, the primary strategy is one already used to discourage students from taking the work of others as their own.

  1. Use plagiarism checkers. Just as we did not know for sure that a fellow student or their sibling did not write their papers, we now fear that we will not be able to discern whether the computer has done its job. Many instructors already rely on plagiarism checking tools. But while a plagiarism detector can’t tell us who wrote a paper if it isn’t in a database of papers to be checked, there is now at least one plagiarism detector dedicated to detecting AI-generated content. If an epidemic of AI work is introduced in school, or even if teachers are convinced of its possibility, there will likely be a proliferation of tools to detect AI typing. Although this may sound promising, I want to add a caveat: In more than ten years of teaching freshman English, I’ve learned that the more I monitor students’ work, the less able I am to be a good teacher. Be careful how much effort you put into this strategy.

The second way: get to know the enemy

The second is the “know your enemy” approach. Artificial intelligence is not going away. It will expand, improve, and become more accurate. Instead of just focusing on discovery, teachers can work around sending AI text in the first place. The strategies of this method are based on designing work that AI cannot do. Here is a representative sample, in order of increasing promise.

  1. Writing in class. Use writing prompts in class. The common notion is that if you watch your students write, they won’t be able to cheat. But in-class writing does not produce every type of writing nor use every skill we want to assess. May discourage writing in favor of a product and may well assess how someone writes under pressure. Although classroom writing can be successfully adopted to measure comprehension and subject knowledge, it does not appear to be the best method for assessing different forms of writing.
  2. writing alternatives. Assign visual organizers or other tasks in place of papers. Over time, AI will likely generate whatever form of personalization we can derive. For now, though, teachers can gauge how well a student’s thesis is supported by ideas, evidence, and arguments, and whether optimal organization is being used. This may lead to presentations in place of written papers, or even collaborative writing sessions during class, if appropriate to the course outcomes.
  3. Topics that avoid the AI ​​command room. Set very specific claims. AI is unlikely to convincingly address written claims of accurate privacy. This is even more true if the claim is about a discussion that occurred in class or some other content that students encountered (guest speakers, peer presentations, field trips, in-class discussions, etc.), which the AI ​​is not aware of. If you ask students to include unique, specific knowledge in their writing, the AI ​​has little chance of including the content you request.
  4. Writing based on human experience. Assign writing that draws on the student’s perspective, experience, and cultural capital. This approach aligns with the diversity, equity, and inclusion model to design writing assignments that can lead to the most significant analysis and synthesis of information. A teacher is as likely to learn from his students’ work as the students. One of the underlying assumptions here is that AI will not produce texts with a resonant personal perspective; But even if AI can replicate this type of writing, the second hypothesis is that a writing task that invites students to share the ways their lives intersect with academia will motivate students to write their own papers.

The last suggestion on this list probably goes back to the “ignorance is bliss” approach, where Hopes Students write their own papers. I see a difference, though, and suspect students will, too, in that the motivation for the two methods is different, with the latter looking for ways to develop and improve the student’s experience with the task.

Method 3: If you can’t beat them, join them

Finally, we have a “if you can’t beat them, join them” approach, where teachers embrace the reality of AI-written content and work with their students to demystify and deconstruct AI-produced textual artifacts. This approach is best suited for classes that have plenty of time to perform AI writing rhetorical analysis, predictions, and evaluation of writing tasks.

  1. rhetorical analysis. Deconstructing the act of typing with AI. Discuss how AI “learns” writing. What assumptions about good writing are revealed when analyzing AI writing? What can’t artificial intelligence do with its writing? Are there write cases in which the AI ​​should be more or less confident? What is the role of humans in creating and proofreading AI scripts?
  2. Peer review. Conduct an AI writing review and/or group discussion. Analyze what he writes. What content does artificial intelligence include? What does not include? How does artificial intelligence organize its writing? What sentence structures do artificial intelligence prefer? Analyze the style in terms of voice, tone, diction, and syntax. Is there a rhythm in the language of artificial intelligence? Can the full rhetorical situation be inferred by analyzing the AI ​​text? How can the text best address the rhetorical situation?
  3. audit. Review AI-generated text. Aside from correcting real-world errors, have students experiment with rearranging the contents of an AI-written piece. Ask the students to expand the paragraphs, combine sentences, add supports, and rewrite the conclusions. Use the AI ​​text as a starting point and opportunity. Students may find it difficult to improve based on “perfection”, but they may also find it easier to revise soulless program writing than their peers.
  4. Class presentations. Provide a comparison/contrast between AI and human typing. Without knowing the author, can students know which text was written by a human and which text was written by an AI? Who writes better? Which writing “looks” better? Compare line by line, thesis statements, audio, organization, evidence and support, arguments and reasoning, overall impact, and persuasive pieces.
  5. revision. Try to get the AI ​​to polish his writing with an emphasis on the rhetorical situation. Have the students make many different shapes of the same vector to adjust the result produced by the AI. Are there limits to how refined the writing can be? Are there trade-offs between one element that are sacrificed when another element is included or enhanced? Ask the students to try to communicate the rhetorical situation by adjusting to the audience, purpose, voice, tone, etc. In the end, is it easier to get the AI ​​to write the text that is exactly appropriate for a given situation or to write it on our own?

There is no wrong or right way to approach the emergence of AI in a writing class. Any coach may use a variety of these strategies. The ideas presented here are not exhaustive, but are presented to enhance thought and add perspective. There is so much more to writing than just composing sentences that I don’t think we need to fear AI being the death knell for composition in education.

Indeed, AI may encourage a brave new exploration of higher-order thinking skills. There are certainly larger conversations about the role of formation courses in higher education – and assessments in all courses – but it can be argued that AI is a tool and that students who learn to use this tool learn a valuable skill.

In ten years, Skynet will probably be writing the five-paragraph articles for everyone and none of this will matter. Or maybe we’re panicking about another 2000. Artificial intelligence will certainly continue to connect the next generation. We can adapt now to make it easier to do this alongside higher education.

Eric Prochaska taught English for over ten years before switching to instructional design. He currently works at Mt. Hood Community College in Oregon, where faculty help design online courses and activities.

Views after: 36

Leave a Comment