29 An AI Workshop for the Overwhelmed and Uninterested

 

As an Instructional Designer I knew the announcement of ChatGPT would impact my job. While I wasn’t certain about the nature of this impact, I recognized the need to understand the pros and cons of AI and how I could harness it to enhance course design. I began this journey by spending time reviewing articles, listening to podcasts, and attending webinars so I could feel somewhat knowledgeable about the uses and pitfalls of AI, specifically ChatGPT and other large language models. The more I experimented with the various tools, the more I wanted to share what they could do to assist instructors with course design. What I didn’t realize was that many instructors did not share my enthusiasm.

My framing of generative AI (GenAI) is similar to handing a caveman a lighter — a tool that reduces workload. Not only less work, but also a tool that boosts student engagement. With this perspective, I searched for ideas on how large language models could assist teaching without the burden of starting over or completely changing lessons and revamping assignments. As an instructional designer many of my consultations with instructors revolve around increasing student engagement throughout their courses. Recommendations might include alternatives to traditional discussion forums or activities to promote reading comprehension. I saw the potential for ChatGPT and other AI tools to add to the list of suggestions I could provide.

In Summer 2023, I was asked to give an informal presentation to a small number of faculty about using AI in their course design. After sharing a few ideas with instructors on how AI could act as a teaching assistant, I discovered that many had not experimented with ChatGPT and were not interested in incorporating it into their courses. A few shared how overwhelmed they felt because AI was changing so fast. Since I had been immersed in the literature about AI, their resistance or lack of knowledge changed my focus from discussing how best to tweak assignments to getting instructors interested in utilizing all AI has to offer. I created a workshop in hopes of piquing their interest and demonstrating that ChatGPT and other GenAI tools could reduce workload and enhance student engagement.

The Workshop Design

The workshop over Zoom began by addressing some common concerns instructors may have regarding AI’s use in the classroom. After sharing my excitement about the possibilities that AI could bring to their courses, I focused on three common prevailing assumptions that seemed to prevent instructors from embracing or even trying AI: concerns related to assignments, academic integrity, and hallucinations.

Assignments

I chose to address assignments in the workshop because many instructors believed any writing assignment needed to be eliminated or handwritten, and they were not ready to make these sweeping changes to their courses. Rather than advocating for the complete overhaul of all written assignments, including discussion posts, I suggested a more measured approach of reviewing and tweaking these assessments. Since AI tools can produce a summary or formulaic essay in seconds, I recommended that written assignments should now include a personal connection to the content. This connection could be a reflection focusing on metacognition. For example, students could provide an explanation about their three biggest takeaways from the chapter, lecture, or video, which enables the instructor to gain a better understanding of the students’ grasp of the material. They could also use course content to make a prediction about what might occur ten or twenty years from now. Another possibility is to require students to include insights from class discussions rather than information already available on the internet. The written response might focus on a small group activity or address their “muddiest point” questions. Although AI tools like ChatGPT may sound conversational, so far they cannot provide personal connections or make predictions based on course materials.

When I am revising a course with an instructor and notice that it consists predominantly of writing assignments, I suggest offering students more choices on how to demonstrate their knowledge and skills. This approach empowers students to showcase their talents in a variety of formats, such as videos, podcasts, and presentations. These alternatives are part of Universal Design for Learning (UDL), a best practice, rather than simply a deterrent to AI. Additionally, students could choose other formats like concept maps, audio files, or infographics to explain their thought process or apply key concepts from the course to real-world or local problems. If a writing assignment still remains the best option, instructors can consider requiring outlines, rough drafts, peer revisions, and a final copy to showcase the full writing process.

Academic Integrity

The workshop also tackled the very real and common concern of academic integrity when using AI tools. This concern frequently arises in my consultations with instructors. Given that both instructors and students often associate the use of GenAI with cheating, this section of the workshop began with strategies to reframe this perception. To prepare students for their future careers, it is essential that both instructors and students understand how to leverage this technology. It cannot be the elephant in the room. Instead, it should be openly discussed in class where we educate students about the benefits and pitfalls of generative AI. By framing our discussions around digital literacy and teaching students to recognize the pros and cons of GenAI, we equip them with valuable knowledge for their futures and eliminate the stigma attached to its use.

An effective approach to initiate class discussion involves being transparent about your own understanding and knowledge of AI. While this field is quickly evolving, no one has all of the answers, but transparency paves the way for more meaningful discussions. When it is time to go beyond a class discussion, a quick and easy micro-lesson demonstrating how AI can be used to brainstorm topics for a project can be a next step. Once the list is generated, students can discuss the ideas that spark their interest or lead them in new directions. If the brainstorming list offers less than optimal choices, it can turn into a lesson on creating prompts that better align with their needs. Creating a brainstorming list together demonstrates the instructor’s willingness to embrace this technology tool.

For those who remain hesitant due to concerns about cheating, I emphasize the importance of fostering a sense of community within the class. When designing courses, I actively seek opportunities to encourage student interaction, both among themselves and with the instructor. Beyond the initial ice breaker activity, how else can students connect? Cultivating a classroom environment where students feel comfortable approaching both the instructor and their peers increases opportunities for learning and reduces incidents of cheating. Building this community involves transparency regarding the applications of AI in the classroom but also establishing clear guidelines that empower students to make informed decisions about its use.

Hallucinations

In the workshop I also addressed the problem of hallucinations. Many instructors have dismissed the use of AI because it can produce fake facts and citations. However, LLMs are also improving everyday. Like other tech tools, the first few versions have bugs that developers eventually fix after receiving feedback. I demonstrated to instructors that hallucinations can be turned into a class activity. Students can analyze an AI-generated paper with in-text citations, facts, quotes, and/or a reference page then evaluate it for accuracy. Do citations match actual people? Are facts accurate and logical? This type of lesson reinforces reviewing work before finalizing and submitting for a grade or feedback. Blind reliance on technology does not make sense, nor does expecting it to create a flawless paper. So examining for hallucinations creates a lesson in digital literacy—a skill we all need to develop.

I also proposed other activities that include critiquing papers for voice and tone. Given that AI lacks the nuance of human expression, students can expand or rewrite an AI-generated paper with an emphasis on integrating a human touch. How does adding a human voice enhance the paper, and why do we need it? The opportunity to discuss these concepts with students helps to foster community and emphasizes the importance of human connection.

Additionally, instructors can leverage AI to address specific writing issues. For instance, if students tend to overuse passive voice, GenAI can generate passages that illustrate this problem, allowing students to practice with how best to correct it. While creating these mini-lessons can be time-consuming for instructors, GenAI can produce examples for students in a matter of seconds. The use of GenAI should not be discounted because of concerns about hallucinations. It helps instructors offer a practical means of imparting valuable lessons in digital literacy while simultaneously saving precious time.

AI as Assistant

In the hopes of intriguing instructors who feel overwhelmed about AI tools, the middle part of the workshop provided ways to reduce instructor workload. Since AI can essentially work as an assistant, I focused on three areas where instructors could benefit from using AI tools: tutoring, brainstorming, and planning.

Tutoring

One of the most valuable aspects of generative AI for instructors is its ability to create examples to illustrate concepts for students who may need more help understanding their coursework. When teaching new content, it can be challenging to find multiple ways to explain or demonstrate complex concepts to students. Asking the GenAI to generate a case study, a few examples, or break down the content into basic steps can save instructors hours of preparation.

When building online courses, it’s challenging to predict which concepts might require more explanation. Therefore, I shared a way for AI to assist students who may struggle with content and need additional support. This help may include offering extra resources and practice materials. Instead of disrupting the flow of class for a few students, instructors can now include a dedicated section or folder within each module that provides additional resources and practice opportunities. These resources can be optional but encouraged for those who need more help. Instructors can create these sections by prompting the GenAI to provide a summary of some of the difficult assigned readings. GenAI can also create a case study with questions and provide answers, enabling students to check their understanding.

As students become more familiar with the capabilities of AI, instructors can encourage its use. Students can prompt the Chatbot to ask questions about a concept or reading, allowing them to be proactive in self-quizzing rather than relying solely on course materials for additional assistance.

Brainstorming

For this part of the workshop, I aimed to reach out to instructors who might be considering making small changes within their courses, providing them with an opportunity to refresh their course through the brainstorming capabilities of AI. When designing a course, instructors often rely on their past experiences, building on what has proven successful. However, certain lessons and projects may begin to feel stale. If grading the same assessments from year to year feels tiring, it might be an indication that it’s time for a change. This change can inject vitality into the course and reignite the instructor’s passion for the subject. When instructors are genuinely excited about a project, it often resonates with their students. Whether the lessons or projects need a small tweak or a complete overhaul, GenAI can provide a plethora of ideas. The brainstorming list might provide twenty possibilities and eighteen of them possibly having been tried or do not align with students’ needs. Nevertheless, the two ideas with potential can help instructors explore methods of assessing students. I have found that this list often acts as a catalyst for my own ideas, taking me in new directions.

The process of course design can be overwhelming and, at times, repetitive. Creating a course map and gathering materials may take months. To reduce the amount of time it takes to design an entire course, GenAI can assist with content curation. It will identify relevant educational resources, such as articles, videos, and case studies. Once the content is chosen, it can provide video captions, article summaries, and discussion questions for students. Additionally, it can be leveraged to monitor relevant sources and provide updates ensuring that courses remain current and relevant. Engaging students effectively is a constant challenge, and seeking input from ChatGPT or another GenAI for brainstorming activities and content can greatly aid instructors in crafting an entire course or a single class. At a minimum, brainstorming with GenAI will stimulate instructors to explore other ways to facilitate learning.

Planning

This part of the workshop was dedicated to harnessing the power of AI in planning class activities. While many of my consultations center around asynchronous course design, our conversations often include various instructional modalities. Regardless of the delivery format, every course should incorporate a deliberate introduction to the week’s topics. As students transition from their out-of-class responsibilities, whether in a face-to-face or synchronous setting, it’s crucial to focus their attention. Following a warm welcome and perhaps a brief overview of the day’s agenda, the focus shifts to the core content of the day’s class. GenAI can provide a thought-provoking question or short video to engage everyone with the current subject.

Moreover, AI can include additional suggestions on how to structure the class effectively. It can generate an organized lesson plan complete with possible activities and a slide deck tailored to the week’s themes and readings. Why spend hours creating a new slide deck or updating the same deck used for years when GenAI can produce a new one in minutes? Instructors only need to use their expertise to conduct a final review and make any necessary revisions.

To further encourage teachers who may still be hesitant about using AI, I shared some of the ways I’ve used it to help with my job. For example, I used ChatGPT to write weekly overviews for a course I was designing. I prompted the tool with weekly objectives, activities, and assessments then asked for a brief summary of about 100 words. Although I had to slightly revise each one, it took minutes rather than hours to complete. I’ve also used AI to create marketing and images for a workshop. The more details I gave, the more nuanced the results.

As the workshop neared the end and considering that instructors had been exposed to numerous ideas, my aim was to address those who might be feeling overwhelmed by the abundance of information. To simplify, I shifted focus to the demonstration of just two AI tools.

Final Push to Create Buy-in

Of the dozens of AI tools available, I chose to eliminate the noise and only demonstrate two that I found particularly helpful. For some, ChatGPT and Bard might be the only tools they’ve heard about, so showing a couple of free, easy to use tools may pique their interest and reduce anxiety. With so many, it was a difficult choice, but with the goal of keeping it simple and free, I demonstrated Teaching Tools and Perplexity. At the time of the workshop, these tools have a free version and can meet a variety of instructional needs.

Perplexity, as a search engine, provides answers to questions with a thorough explanation. Along with the answer, it also offers a list of the resources used to generate the response to the question. Clicking on the resources allows for the opportunity to explore other ideas and determine credibility, relevancy, and currency. Each time I’ve tried it, I received a minimum of four resources, which I reviewed, leading me to discover more resources. With a free account, the search engine also maintains a user’s history, enabling the return to previously asked questions.

Teaching Tools provides everything needed to plan an entire course or just one class. It also has a short 2-minute video explaining the brainstorming tool that I showed during the workshop. Essentially, the instructor can give the title of the course and then describe it as if explaining it to a colleague. From there, users can choose learning objectives, lecture topics, discussion questions and/or case studies. I really liked the responses I received when I asked it to generate discussion questions based on specific learning goals and topics. It provided me with 10 questions and possible answers. Some of the answers included examples pertaining to the opioid crisis and pandemic, which I found to be relevant and more current than I expected.

Since these tools are easy to use, and I only demonstrated two, it gave instructors who may have come to the workshop feeling overwhelmed because of the many AI choices, a place to start.

Workshop Follow-up and Next Steps

The response to the workshop has been generally positive, with participants showing interest in trying out the tools introduced. Some have started experimenting with integrating AI into their teaching, while a few expressed concerns about students relying too much on AI for assignments and not experiencing the struggles associated with learning. During the workshop, some of the participants remained quiet. Rather than interpreting this as a negative, I see it as another part of the change process. As with any other new tool, instructors are taking the time to process the possibilities and consider how to incorporate AI into their individual teaching practices. The process of change requires time to think and experiment.

In the days after the workshop, some instructors shared their reimagined assessments where they attempted to “AI-proof” writing assignments by focusing on more reflective and personal responses. They used GenAI to help generate reflection questions. Their small changes represent a first step in recognizing the potential benefits of AI in education. I’ve also had in-depth conversations with instructors exploring the use of AI for grading, looking into how it can streamline assessments and provide personalized feedback while respecting student privacy. For my next workshop, I plan to add examples of how AI can help with the grading process.

My enthusiasm for GenAI has only grown. I still encounter instructors who are wary of AI tools and others who don’t know where to begin, but I remain committed to guiding them through the gradual process of incorporating AI into their course design. I will continue offering workshops and brief demonstrations of specific AI tools with a focus on creating understanding and showing practical ways to use AI for saving time and improving student engagement.

As AI technology evolves at a record pace, both instructors and students need time to explore its applications for their specific needs. I’ll continue to help educators navigate this exploration, providing support and insights to make the integration of AI into course design a smoother process.