Just as calculators, personal computers, and smartphones have threatened students learning math skills, AI (artificial intelligence) is emerging as a new tool that seeks to undermine the use of essay tasks to assess student learning. It seems that.
In November 2022, a tool called ChatGPT made headlines for its ability to “write” any content. As an instructional designer, facing a robot capable of writing student papers, wondering what opportunities they have, from a faculty member worried the sky might fall I heard it right away.
After some thought, I’ve come to believe that in the long run, worrying about students potentially cheating with AI isn’t the most productive issue to focus on. A better question is how best to teach students, even in the age of AI. Below are his three ways to design writing assignments in the face of AI invasion.
Method 1: Ignorance is bliss
In extreme reactions, there is the approach that “ignorance is bliss” and “resistance is futile.” These attitudes are lumped together because both attitudes prefer to avoid the core issue. In the former case, the instructor may simply be unaware that the student is now able to enter writing prompts on her website and copy the generated responses into a document for submission. In the latter case, the instructor may be aware of her AI’s writing prowess, but metaphorically arm herself with the overwhelming notion that she can no longer tell if a student has written a submitted paper. may throw.
Worst of all, teachers with this mindset will give themselves up to AI-written grading and hope that most students are still writing their own papers and learning from the feedback. Student Writing For an instructor who uses assessments to help her improve her skills, responding to something a student has not written is a waste of time. Also, these students invest little in reviewing feedback.
For instructors who recognize the ability of AI to produce reports but feel unprepared to tackle robots head-on, a key strategy is for students to disguise the work of others as their own. already used to prevent
- Hire a plagiarism checker. Just as I couldn’t know for sure that a student’s classmates or siblings hadn’t written a paper, I worry that the computer wouldn’t be able to discern if it had done its job. Instructors already rely on plagiarism checkers. However, although a plagiarism detector cannot tell who wrote a paper if it is not in the database of papers it checks, there is at least one plagiarism detector dedicated to stealing AI-generated content. If an AI work epidemic were to be filed in schools, or even if instructors were convinced of its potential, there could be a proliferation of tools to detect AI writing. , I’d like to add a note of caution: In my over ten years of teaching freshman English, I’ve learned that the more you police your students’ work, the less energy it takes to be a good teacher. Be careful how much effort you put into this strategy.
Method 2: Know Your Enemy
The second is the “know your enemy” approach. AI will never go away. It will expand, it will improve, it will become more subtle. Instructors can avoid submitting AI text in the first place, rather than focusing solely on detection. The strategy of this method relies on designing work that AI cannot do. Below is a representative sample, listed in descending order of promise.
- writing during class. Use in-class writing prompts. The general notion is that you can’t cheat if you watch students write. However, in-class writing does not generate all types of writing or involve all skills you want to assess. In-class writing can be well employed to measure comprehension and subject knowledge, but it is not the best way to assess different forms of writing. Doesn’t seem to be the way.
- Write an alternative. Assign visual organizers or other assignments instead of documents. Eventually AI will generate any form of assignment we can think of. For now, however, instructors can measure how well a student’s paper is supported by ideas, evidence, and arguments, and whether the optimal structure is used. If so, you may do a presentation instead of a paper or have a co-writing session during class.
- Topics to avoid AI wheelhouses. Assign very specific prompts. AI is less likely to convincingly respond to well-written prompts. This is even more true if the prompt is related to a discussion that occurred in class or other content encountered by students that the AI is unaware of (e.g. guest speakers, peer presentations, excursions, in-class discussions, etc.) . If you ask students to include their own specific knowledge in their writing, AI is unlikely to include the required content.
- Write based on human experience. Assign sentences that depend on the student’s perspective, experience, and cultural capital. This approach is consistent with diversity, equity, and inclusion models that design writing assignments that are likely to result in the most meaningful analysis and synthesis of information. Instructors are just as likely to learn from their students’ work as they are from their students. One of the underlying assumptions here is that AI does not generate text with a resonating personal point of view. But even if AI can replicate this kind of writing, the second premise is that writing challenges that encourage students to share how their lives and academia intersect motivate students to write their papers. is to give
Perhaps the last suggestion on this list is that the instructor hope Students write their papers. But I see a difference. I also think that the students, as well, differ in that the motivations behind the two methods are different, with the latter looking for ways to evolve and improve the student’s experience of the assignment.
Method 3: Join if you can’t win
Finally, there is the “if you can’t beat them, join in” approach where instructors embrace the reality of AI-generated content and work with students to demystify and deconstruct AI-generated text artifacts. This approach works best for classes that have plenty of time to perform rhetorical analysis of AI writing and expectations and assessment of writing tasks.
- rhetorical analysis. Deconstructing the act of AI lighting itself. Discuss how the AI ”learns” to write. What assumptions about good writing do your analysis of AI writing reveal? What is AI unable to do with writing? What is the human role in text generation and proofreading?
- peer review. Conduct peer reviews and class discussions on AI writing. Analyze what is written. What content does AI include? What does it not include? How does AI organize sentences? What sentence structure does AI prefer? It analyzes your style in terms of voice, tone, phrasing and syntax. Does AI language have rhythm? Can we infer complete rhetorical situations by analyzing AI texts? How well can texts handle rhetorical situations?
- revision. Fix AI-generated text. Have students experiment with rearranging the content of AI-written work, rather than just correcting factual errors. Have students extend paragraphs, join sentences, add supplements, and rewrite conclusions. Use AI text as a starting point and as an opportunity. A student may find it difficult to improve to “perfect”, but may find it easier to fix the soulless program writing than his peers.
- class presentation. Present a comparison/contrast between AI and human writing. Without knowing the authors, can a student know which texts were written by a human and which were written by her AI?Who writes well? Which style of writing “sounds” better? Line by line, compare the paper’s statement, voice, construction, evidence and support, argument and logic, overall impact, and partly persuasiveness.
- sophistication. Focus on rhetorical situations and let the AI refine your sentences. Have students create several variations of the same prompt to fine-tune the AI-generated results. Are there limits to the extent to which the writing can be improved? Are there trade-offs where one element is sacrificed when another is included or enhanced? to fit the rhetorical situation. Ultimately, is it easier to have an AI create text that is perfectly suited to your particular situation, or to create it yourself?
There is no wrong or right way to deal with the emergence of AI in your lighting class. Any instructor may employ a variety of these strategies. The ideas presented here are not exhaustive, but are provided to encourage thought and add perspective. Writing does much more than the act of creating, so I don’t think we need to fear that AI will be the death knell of creation in education.
In fact, AI could facilitate brave new explorations of higher-order thinking skills. While there is no doubt that the role of writing courses in higher education, and the role of assessment in all courses, needs greater debate, AI is a tool and students learning how to use it will acquire valuable skills. You can argue that you are learning.
Ten years from now Skynet may be writing everyone’s five-paragraph essay, but that doesn’t matter. Or maybe he’s panicking about Y2K. AI will definitely plug out in the next generation as well. We can make adjustments now to advance it hand in hand with higher education.
Eric Prochaska taught English for over ten years before shifting his focus to instructional design. He currently works at a community college in Mount Hood, Oregon, where he helps faculty design online courses and activities.
Post view: 239
