Lunch and Learn: Generative AI Uses in the Classroom

On Tuesday, April 23rd, the Center for Teaching Excellence and Innovation (CTEI) hosted a Lunch and Learn on Generative AI Uses in the Classroom. Faculty panelists included Louis Hyman, Dorothy Ross Professor of Political Economy in History and Professor at the SNF Agora Institute, Jeffrey Gray, Professor of Chemical and Biomolecular Engineering in the Whiting School, and Brian Klaas, Assistant Director for Technology and instructor at the Bloomberg School of Public Health. Caroline Egan, Teaching Academy Program Manager, moderated the discussion.  

Louis Hyman began the presentation by reminding the audience what large language models (LLMs) like ChatGPT can and cannot do. For example, ChatGPT does not “know” anything and is incapable of reasoning. It generates text that it predicts will best answer the prompt it was given, based on how it was trained. In addition to his course work, Hyman mentioned several tasks he uses ChatGPT to assist with, including text summarization, writing complicated Excel formulas, writing and editing drafts, making PowerPoint tables, and turning image files in the right direction.

In Hyman’s course, AI and Data Methods in History, students are introduced to a variety of tools (e.g., Google Sheets, ChatGPT, Python) that help them analyze and think critically about historical data. Hyman described how students used primers from LinkedIn Learning as well as Generative AI prompts to increase their technical skills which enabled them to take a deeper dive into data analysis. For example, while it would have been too complicated for most students to write code on their own, they learned how to prompt ChatGPT to write code for them.  By the end of the semester, students used application programming interface (API) calls to send data to Google, used OpenAI to clean up historical documents and images presented using optical character recognition (OCR), and used ChatGPT and Python to plot and map historical data.Two maps of 1850 New England showing the number of congregational churches and the value of congregational property. Data points plotted by students using AI.

Hyman noted that one of the most challenging parts of the course was convincing students that it was OK to use ChatGPT, that they were not cheating.  Another challenge was that many students lacked basic computer literacy skills, therefore, getting everyone up to speed took some time. There was also not one shared computer structure/platform. The successes of the course include students’ ability to use libraries and APIs to make arguments in their data analysis, apply statistical analysis of the data, and ask historical questions about the results they were seeing in the data.

Jeff Gray continued by describing his Computational Protein Structure Prediction and Design course that he has taught for over 18 years. In this course, students use molecular visualization and prediction tools like PyRosetta, an interactive Python-based interface that allows them to design custom molecular modeling algorithms. Recently, Gray has introduced open-sourced AI tools into the curriculum (AlphaFold and RoseTTAFold), which predict 3D models of protein structures.

Example of protein folding using AlphaFold.

One of the challenges Gray mentioned was the diversity of student academic backgrounds. There were students from engineering, biology, bioinformatics, computer science, and applied math, among others. To accommodate this challenge, Gray used specifications grading, a grading method in which students are graded pass/fail on individual assessments that align directly with learning goals. In Gray’s class, students were presented with a bundle of problem sets categorized at various difficulty levels. Students selected which ones they wanted to complete and had the option of resubmitting them a second time for full credit. Gray is undecided about using this method going forward, noting that half of the students ended up dropping the course when they tried to complete all of the problems instead of just a few, and found the workload too heavy.  Another challenge was how to balance the fundamental depth of the subject matter versus application.  To address this, Gray structured the twice weekly class with a lecture on one day and a hands-on workshop the other day, which seemed to work well.

Brian Klaas teaches a one credit pass/fail course called Using Generative AI to Improve Public Health. The goal of this course is to allow students to explore AI tools, gain a basic understanding of how they work, and then apply them to their academic work and research. In addition to using the tools, students discussed the possible harms in Generative AI, such as confabulations, biases, etc., the impact of these tools in Public Health research, and future concerns such as the impact on the environment and copyright law. Klaas shared his syllabus statement regarding the usage of AI tools in class, something he strongly recommends all faculty share with their students 

Hands-on assignments included various ways of using Generative AI. In one assignment, students were asked to write a summary of a journal article and then have GenAI write a summary of the same article geared towards different audiences (academics vs. high school students). Students were then asked to analyze the differences between the summaries.Sample instagram post created using AI showing people from different cultures dressed as medical professionals. For another assignment, students were asked to pick from a set of topics and use Generative AI to teach them about the selected topic, noting any confabulations or biases present. They then asked GenAI to create a five-question quiz on the topic and take the quiz. A final assignment was to create an Instagram post on the same topic including a single image and a few sentences explaining the topic to a lay audience. All assignments included a reflection piece which often required peer review.

Lessons learned: Students loved the interdisciplinary approach to the course, confabulations reinforce core data research skills, and learning from each other is key.

The discussion continued with questions from the audience: 

Q: What would you recommend to an instructor who is considering implementing GenAI in the classroom? How do they start thinking about GenAI?
JG: Jupyter notebooks are pretty easy to use. I think students should just give it a try.
LH: I recommend showing students what ”bad” examples look like. The truth is, we can still write better than computers. Use AI to draft papers and then use it as an editing tool – it’s very good as an editing tool. Students can learn a lot from that.
BK : I recommend having students experiment and see where the strengths lie, get an overall awareness of it. Reflect on that process, see what went well, not so well. Feed in an assignment and see what happens. Use a rubric to evaluate the assignment. Put a transcript in and ask it to create a quiz on that information. It can save you some time.

Q for Brian Klaas: What version of GPT were you using?
BK: Any of them – I didn’t prescribe specific tools or versions. We have students all over the world, so they used whatever they had. ChatGPT, Claude, MidJourney, etc. I let the students decide and allowed them to compare differences.

Q for Jeff Gray: Regrading the number of students who dropped, is the aim of the course to have as many students as possible, or a group who is wholly into it?
JG: I don’t know, I’m struggling with this. I want to invite all students but also need to be able to dig into the math and material. It feels like we just scratched the surface. Maybe offering an intersession course to learn the tools before they take this class would be helpful. There is no standard curriculum yet for AI. Where to begin…we’re all over the map as far as what should be included in the curriculum.
LH: I guess it depends on what your goals are. Students are good at “plug and chug,” but bad at asking questions like, “what does this mean?”
BK: We didn’t get to cover everything, either – there is not enough time in a one credit class. There are just so many things to cover.

Q: What advice do you have for faculty who are not computer scientists? Where should we start learning? What should we teach students?
LH: You can ask it to teach you Python, or how to do an API call. It’s amazing at this. I don’t know coding as well as others, but it helps. Just start asking it [GenAI]. Trust it for teaching something like getting Pytorch running on your PC. Encourage students to be curious and just start prompting it.
BK: If you’re not interested in Jupyter notebooks, or some of the more complicated functions, you can use these tools without dealing in data science. It can do other things. It’s about figuring out how to use it to save time, for ideation, for brainstorming.
JG: I have to push back – what if I want to know about what’s going on in Palestine and Israel? I don’t know what I don’t know. How do I know what it’s telling me is correct?
LH: I don’t use it for history – but where is the line of what it’s good and not good at?
BK: I would use it for task lists, areas to explore further, but remember that it has no concept of truth. If you are someone who knows something about the topic, it does get you over the hurdles.
JG: You have to be an expert in the area to rely on it.
LH: Students at the end of my course made so much progress in coding. It depends on what you ask it to do – protein folding is very different than history that already happened.

Q: How can we address concerns with fairness and bias with these tools in teaching?
BK: Give students foundational knowledge about how the tools work. Understand that these are prediction machines that make stuff up. There have been studies done that show how biased they are, with simple prompts. Tell students to experiment – they will learn from this. I suggest working this in as a discussion or some practice for themselves.

Q: Students have learned to ask questions better – would you rather be living now with these tools, or without them?
JG: Students are brainstorming better. They are using more data and more statistics.
BK: AI requires exploration and play to get good responses. It really takes time to learn how to prompt well. You have to keep trying. Culturally, our students are optimized for finding the “right answer;” AI programs us to think that there are multiple answers. There is no one right answer for how to get there.
LH: Using AI is just a different process to get there. It’s different than what we had to do in college. It was hard to use computers because many of us had to play with them to get things to work. Now it all works beautifully with smart phones. Students today aren’t comfortable experimenting. How do we move from memorization to asking questions? It’s very important to me that students have this experience. It’s uncomfortable to be free and questioning, and then go back to the data. How do we reconcile this?

JG: What age is appropriate to introduce AI to kids?
LH: Students don’t read and write as much as they used to. I’m not sure about the balance.
Guest: I work with middle and high school teachers. Middle school is a great time to introduce AI. Middle school kids are already good at taking information in and figuring out what it means. Teachers need time to learn the tools before introducing it to students, including how the tools can be biased, etc.

Q: How can we encourage creative uses of AI?
BK: Ethan Mollick is a good person to follow regarding creative uses of AI in education and what frameworks are out there. To encourage creativity, the more we expose AI to students, the better. They need to play and experiment. We need to teach them to push through and figure things out.
LH: AI enables all of us to do things now that weren’t possible. We need to remember it’s an augment to what we do, not a substitute for our work.

Resources:
Hyman slides
Gray slides
Klaas slides

Amy Brusini, Senior Instructional Designer
Center for Teaching Excellence and Innovation
 

Image source: Lunch and Learn logo, Hyman, Gray, and Klaas presentation slides, Unsplash

Lunch and Learn: Generative AI – Teaching Uses, Learning Curves, and Classroom Guidelines

On Tuesday, October 3rd, the Center for Teaching Excellence and Innovation (CTEI) hosted its first Lunch and Learn of the academic year, a panel discussion titled, “Generative AI: Teaching Uses, Learning Curves, and Classroom Guidelines.” The three panelists included Jun Fang, Assistant Director of the Instructional Design and Technology Team in the Carey Business School, Carly Schnitzler, KSAS instructor in the University Writing Program, and Sean Tackett, Associate Professor in the School of Medicine.  The discussion was moderated by Caroline Egan, project manager in the CTEI. Mike Reese, director of the CTEI, also helped to facilitate the event. 

The panelists began by introducing themselves and then describing their experiences with generative AI. Jun Fang loves new technology and has been experimenting with AI since its inception. He noticed the faculty that he works with generally fall into two categories when it comes to using AI: some are quite concerned about students using it to cheat and are not ready to use it, while others see a great deal of potential and are very excited to use it in the classroom.  In speaking with colleagues from across the institution, Fang quickly realized these are common sentiments expressed by faculty in all JHU divisions. This motivated him to lead an effort to create a set of AI guidelines specifically geared toward faculty. The document contains a number of strategies for using AI including: designing engaging course activities, providing feedback for students on their assignments, and redesigning course assessments. The section on redesigning course assessments uses two approaches: the “avoidance approach,” which involves deliberately designing assessments without AI, and the “activation approach,” which intentionally integrates AI tools into the curriculum. The document includes specific examples of many of the strategies mentioned as well as links to widely used generative AI tools. 

Fang described a recent scenario in which a faculty member was concerned that students were using ChatGPT to generate answers to online discussion board questions.  To mitigate this situation, Fang suggested the faculty member revise the questions so that they were tied to a specific reading or perhaps to a topic generated in one of his online synchronous class sessions.  Another suggestion was to have students submit two answers for each question – one original answer and one generated by ChatGPT – and then have the students compare the two answers.  The faculty member was not comfortable with either of these suggestions and ended up making the discussion more of a synchronous activity, rather than asynchronous.  Fang acknowledged that everyone has a different comfort level with using AI and that one approach is not necessarily better than another.     

Carly Schnitzler currently teaches two introductory writing courses to undergraduates and is very open to using generative AI in her classroom.  At the start of the semester, she asked students to fill out an intake survey which included questions about previous writing experiences and any technologies used, including generative AI. She found that students were reluctant to admit that they had used these technologies, such as ChatGPT, for anything other than ‘novelty’ purposes because they associated these tools with cheating. After seeing the results of the survey, Schnitzler thought it would be beneficial for students to explore the potential use of generative AI in class. She asked students to do an assignment where they had to create standards of conduct in a first year writing class, which included discussing their expectations of the course, the instructor, their peers, and how AI would fit in among these expectations. The class came up with three standards: 

  1. AI tools should support (and not distract from) the goals of the class, such as critical thinking, analytical skills, developing a personal voice, etc.  
  2. AI tools can be used for certain parts of the writing process, such as brainstorming, revising, or editing, but students must disclose that AI tools were used. 
  3. If there appears to be an over-use or over-reliance on AI tools, a discussion will take place to address the situation rather than disciplinary action. (Schnitzler wants students to feel safe exploring the tools without fear of repercussion.) 

This assignment comes from an open collection of cross-disciplinary assignments that use text generation technologies, mostly in a writing context. TextGenEd: Teaching with Text Generation Technologies, co-edited by Schnitzler, consists of freely accessible assignments submitted by scholars from across the nation. Assignments are divided into categories, such as AI literacy, rhetorical engagements, professional writing, creative explorations, and ethical considerations. Most are designed so that the technologies used are explored by students and instructors together, requiring very little ‘expert’ technological skills.  Schnitzler noted that there is a call for new submissions twice each year and encouraged instructors to consider submitting their own assignments that use text generation AI.

Sean Tackett was initially fearful of ChatGPT when it was released last year. Reading article after article stating how generative AI was going to “take over” pushed him to learn as much as he could about this new technology. He began experimenting with it and initially did not find it easy to use or even necessarily useful in his work with medical school faculty. However, he and some colleagues recognized potential in these tools and ended up applying for and receiving a JHU DELTA grant to find ways they could apply generative AI to faculty development in the medical school. Tackett described how they are experimenting with generative AI in a curriculum development course that he teaches to the med school faculty. For example, one of the tasks is for faculty to learn to write learning objectives, so they’ve been developing prompts that can be used to specifically critique learning objectives. Another example is developing prompts to critique writing. Most of Tackett’s students are medical professionals who do not have a lot of time to learn new technologies, so his team is continually trying to refine prompts in these systems to make them as useful and efficient as possible. Despite being so busy, Tackett noted the faculty are generally enthusiastic about having the opportunity to use these tools.     

The discussion continued with a question and answer session with audience members: 

Q: How do we transfer and integrate this knowledge with teaching assistants who help manage the larger sized classes? What about grading?
ST: I would advocate for the potential of AI to replace a TA in terms of grading, but not in terms of a TA having a meaningful dialogue with a student. 
JF: Generative AI tools can be used to provide valuable feedback on assessments. There are a lot of tools out there to help make grading easier for your TAs, but AI can be used for the feedback piece. 

Q: How might professors provide guidelines to students to use generative AI to help them study better for difficult and complex topics?
MR: One possibility is to generate quiz questions – and then have students follow up by checking the work of these quizzes that have been generated.
CS: Using a ChatGPT or other text generation tool as a reading comprehension aid is something that has been useful for non-native English speakers. For example, adding a paragraph from an academic article into ChatGPT and asking what this means in plain language can be helpful.

CE: This gets to what I call ‘prompt literacy,’ which is designing better prompts to give you better answers. There is a very good series about this on Youtube from the University of Pennsylvania.
Sean, what have you experienced with prompting right now, in terms of challenges and opportunities?
ST: We’re trying to put together advice on how to better prompt the system to get more refined and accurate answers. After a few iterations of prompting the system, we refine the prompt and put it into a template for our faculty, leaving a few ‘blanks’ for them to fill in with their specific variables. The faculty are experts in their subject areas, so they can tell if the output is accurate or not. We’re in the process of collecting their output, to put together best practices about what works, what does not work.  

CE: What would you all like to see in terms of guidelines and best practices for AI on a web page geared towards using AI in the classroom?
Guest: And along those lines, how to we move forward with assigning research projects, knowing that these tools are available for students?
ST: I think it could be useful for students to learn research skills. They could use the tools to research something, then critique the results and explain how they verified those results. It can also be useful for generating ideas and brainstorming. Another thought is that there are a number of domain specific generative AI databases, such as Open Evidence which is useful in the medical field.  
CS: To Sean’s point, I think a comparative approach is useful with these tools. The tools are very good at pattern matching genre conventions, so doing comparative work within a genre could be useful.
JF: I think ChatGPT and other generative AI tools can be useful for different parts of the research process, such as brainstorming, structure, and editing. But not for something like providing or validating evidence.  

Q: As a grad student, I’m wondering how the presence of AI might force us to refine the types of questions and evaluations that we give our students. Are there ways to engineer our own questions so that the shift of the question is changed to avoid the problem [of having to refine and update the question] in the first place?
CS: There is an assignment in our collection that talks about bringing an assignment from past to present. Again, thinking in terms of a comparative approach, ask ChatGPT the question, and then ask your students the same question and see how they compare, if there are any patterns.  I think it can be helpful to think of ChatGPT as adding another voice to the room.
JF: We have a section in the guidelines on how to redesign assessment to cope with generative AI related issues. We suggest two approaches: the avoidance approach and the activation approach. The avoidance approach is for faculty who are not yet comfortable using this technology and want to avoid having students use it.  One example of this approach is for faculty to rework their assignments to focus on a higher level of learning, such as creativity or analysis, which will hopefully reduce or eliminate the opportunity for students to use AI tools. The activation approach encourages faculty to proactively integrate AI tools into the assessment process. One example of this approach I mentioned earlier is when I suggested to a faculty member to rework their discussion board questions to allow students to submit two versions of the answers, one created by them and the other by ChatGPT, and then analyze the results. 

Q: What is the ultimate goal of education? We may have different goals for different schools. Also, AI may bridge people from different social backgrounds. In China, where I grew up, the ability to read or write strongly depends on the social status of the family you come from. So there is some discomfort using it in the classroom.
CS: I feel some discomfort also, and that’s what led to the development of the guidelines in my classroom. I posed a similar question to my students: if we have these tools that can allegedly write for us, what is the point of taking a writing class?  They responded by saying things like, “writing helps to develop critical thinking and analytical skills,” to which I added, “being here is an investment in yourself as a student, a scholar, and a thinker.” I think asking students to articulate the value of the education that they want to get is really helpful in determining guidelines for AI.
ST: Going to school and getting an education is an investment of your time. You pay now so you can be paid later. But it’s not as transactional as that. AI is already in the work environment and will become more prevalent. If we’re not preparing students to succeed in the work environment, we are doing them a disservice. We teach students to apply generative AI in their classes so they are prepared to use it in the workforce.
JF: In the business school, everything is market driven. I think education can fit into that framework as well. We’re trying to provide graduates with the confidence they need to finish the work and meet the market’s need. We know that generative AI tools have really changed the world and they’re starting to emerge in every part of our life. We need to train students to realize that ChatGPT might be part of their education, part of life in the future, and part of the work in the future as well. There are things AI can help us do, but there are still fundamentals that students need to learn. One example is calculators: we still need to learn from the beginning that 1 + 1 = 2. 
CE: This question also reminded me of asking your students, what is the ultimate purpose of a research paper? Where do they think ChatGPT should fit into the research process?  

Q: I work at the library and we’re getting lots of questions about how to detect if students are using AI. And also, how do you determine if students are relying too heavily on AI?
JF: We also get this question from our faculty. The most used detection tool right now is Turnitin, which is embedded in Canvas. But the level of accuracy is not reliable. We encourage faculty to always validate before accepting the results.  For faculty who are actively using AI in the classroom, we also encourage them to provide clear guidance and expectations to students on how they are allowed to use it.  This may make it a little easier to determine if they are using it correctly or not.
MR: There are some other tools out there, such a GPTZero, ZeroGPT, but to Jun’s point, the difficult thing is that it’s different than plagiarism detection which says this is copied, and here’s the source. These tools say there’s a probability that part of this was taken, but you can’t point to a direct source. It’s up to instructors whether or not to use these tools, but consider using them to facilitate a conversation with students. In my own classes if I suspect academic misconduct, I usually start by asking them to explain, talk to me about what is happening before I make accusations. With these tools, there tends to be no hard evidence, just probabilities that something may have happened.  This is definitely an area we’re all still learning about.
Guest: I was just thinking that having a conversation with students about why they are turning to the tool in the first place might prevent misconduct.  Instead of sending them to an academic misconduct committee, we could have these conversations, like Carly mentioned. Making students aware of the limitations of the tool could also be helpful.
CS: Yes, I say that in our guidelines that I’m prioritizing conferences with students over immediate disciplinary action. I try to pre-empt anxiety students might feel around using these tools. Designing your assignments in a way that reduces anxiety is also helpful. For example, I tend to design assignments that build on one another throughout the semester in smaller bits, rather than one giant chunk all at once.  

Q: Is there any discussion around combining AI with teaching, such as generating personalized explanations of a topic? Students will have different levels of expertise and comfort with different topics.
ST: We’re trying to do this, to create a teaching aid for the future. We’re planning to use it to create assessment items.  

Amy Brusini, Senior Instructional Designer
Center for Teaching Excellence and Innovation
 

Image Source: Pixabay, Unsplash

 

Adapting to AI in the Classroom for Time-Strapped Instructors

In the past few months, we have spoken to many instructors – faculty, graduate students, even undergraduate teaching assistants –  who are doing very interesting things with artificial intelligence tools in their classes this coming fall. Some are writing grants to support research into classroom uses of AI, some are designing interactive online modules to help teach about the ethics of AI, and some are integrating AI tools into their instructional activities.

This blog post is for another instructor population: those that have not had the time or capacity to redevelop their courses, their assessments, or their activities to accommodate an AI world. “Redesigning assessments with AI in mind” might be the 20th item on a long list of to-dos for the coming semester. Adapting to new technologies that could change the classroom experience – and AI is certainly one of them – seems like an overwhelming task. Classes start in one week, and wrestling with the teaching and learning opportunities and challenges of artificial intelligence may not be an achievable goal.

However, there are some concrete steps and curated resources to take into account in terms of AI when planning and teaching your courses.

Recommendations for Starting with AI

Here are six recommendations (and one extra credit assignment). Following all of these suggestions will put you on good footing with the learning curve associated with AI in the classroom, but even doing one or two is a good way to start.

  1. Experiment with ChatGPT and other AI tools. Just get in there and start using them and see what they produce. In an article for the Chronicle of Higher Education, one writer said, “I started by reminding myself, anytime I was about to Google something, to ask ChatGPT.”[1] ChatGPT-ing (or using Google Bard) instead of Google-ing is a good on-ramp to AI usage. You may even find them useful to you as an instructor. Here are four basic generative AI models to start with along with prompt suggestions:
    1. ChatGPT – The first (and by some reports, still the most accurate) text-based generative AI. Prompt suggestion: Ask a basic question about teaching, e.g., “How can I grade exams more efficiently?” or “How can I provide written feedback more efficiently?”
    2. Google BardLess text-heavy than ChatGPT; potentially geared towards more logic-based questions, e.g., “How do I create a website in WordPress?”
    3. Microsoft BingAble to generate images as well as text and simultaneously harness the power of a search engine. Potential question: “Name the characteristics of neo-classical architecture and provide an example.”
    4. Fotor.com Image-generator AI. Potential question: “Provide an illustration for my chemistry class syllabus.”
  2. Run your assignments through an AI tool. This will help benchmark possible AI-generated responses to your assignments. More sophisticated AI users will engage in prompt engineering that could make uncited or incorrect usage of AI harder to detect, but getting at least one example of an AI response is helpful. It will not only provide a sightline into possible academic integrity issues but also point to whether your assignment may need to be revised or redeveloped, which could include integrating AI itself. Derek Bruff, a writer and higher education consultant, provides good guidance on assessment design in light of AI:
    1. Why does this assignment make sense for this course?
    2. What are specific learning objectives for this assignment?
    3. How might students use AI tools while working on this assignment?
    4. How might AI undercut the goals of this assignment? How could you mitigate this?
    5. How might AI enhance the assignment? Where would students need help figuring that out?
    6. Focus on the process. How could you make the assignment more meaningful for students or support them more in the work? [2]
  3. Add an AI policy to your syllabus. This may require doing some or all of the recommendations above, but even if you do not have the capacity to take a deep dive into AI tools before courses start, it is a good idea to take a stab at a policy, even if it is brief. As mentioned above, you will be adapting this policy fairly quickly. The sooner you develop a benchmark policy and determine what works and what does not, the better. Lance Eaton, a doctoral student in higher education at the University of Massachusetts at Boston, has crowdsourced a Google Document with many helpful examples of AI policies for syllabi. This is an excellent place to start.
  4. Determine your academic integrity policy for AI. This may be part of your general AI policy or it could be separate. Regardless, this will probably be V.1 of your academic integrity policy, but again, starting now will put you in a good position to iterate as needed. To start, review Academic Integrity Policies for Johns Hopkins Schools. Lance Eaton’s Google Document (above) has many examples of AI policies that include academic integrity statements.
  5. Teach your students how to cite AI tools. This information could be incorporated into a syllabus policy and/or academic integrity policy, but correct citation – at least according to August 2023 recommendations of these style guides – is step number one. Making your students aware that they need to cite uses of AI tools and giving them the tools for doing that will (hopefully) incentivize compliance with your academic integrity policies.
    1. APA Citation Guidance – ChatGPT
    2. MLA Citation Guidance – Generative AI
    3. Chicago Style Citation Guidance – ChatGPT
    4. Johns Hopkins Library Guide on Citation
  6. Talk to your local center for teaching and learning. All Hopkins Schools have teaching and learning centers, some have been publishing guidance on how to teach and learn with artificial intelligence tools, and many have been considering the possible consequences of AI in the classroom. Here’s a list of teaching and learning centers at Hopkins, and here are two rich resources developed by two CTLs at Hopkins:
    1. Teaching & Learning in the ChatGPT Era. This website was created by the Center for Learning Design & Technology at the Whiting School of Engineering. It provides a great overview on generative AI as well as providing guidance on academic integrity questions, student use of AI, and assessment design with AI. Kelly Orr, Nathan Graham, Olysha Magruder, Mel Rizzuto, and Edward Queen of the CLDT all contributed to the website as did adjunct faculty David Porter.
    2. Johns Hopkins University Generative AI Tool Implementation Guidance and Best Practices. Jun Fang, Assistant Director in Teaching & Learning@Carey in the Carey School of Business led the development of this resource with contributions from representatives at other schools and teaching and learning centers at Hopkins. This guide provides substantial guidance on using generative AI to design engaging course activities, provide assignment feedback, and gives a list of AI tools for higher education.

Extra credit assignment for those with a little more capacity:

  1. Learn a little about prompt engineering. Prompt engineering is developing and refining questions and statements for AI models such that they generate results with the desired specificity, tone, length, citations, etc. This will give you a sightline into AI capacities beyond a simple one-time command (e.g., “Compare and contrast models of femininity in Shakespeare’s Much Ado About Nothing and Taming of the Shrew”) which may yield an overly broad answer that lacks specificity and nuance. Prompt engineering will also help you learn to direct and guide AI models and not just react to them. For a useful beginner’s guide to prompt engineering, check out the brief video on prompting AI from Wharton School instructors.

Why You Should Do This

Here is why you should take the (small) leap: Artificial intelligence will change the way we teach and learn. The internet did this, email did this, and so will AI. Taking small steps to acculturate to this new reality is the best way to build the flexibility needed to successfully teach and learn with AI – and, very importantly, teach your students how to teach and learn with AI. Here are more reasons to begin to shift your behavior:

  • You can start small. Take this semester as an opportunity to begin to build your AI teaching and learning skills. You do not have to overhaul your syllabi or classroom activities to accommodate AI; you just have to begin to think through the implications of teaching in a world where AI tools are easily available and could pass your homework assignments. Ask yourself how you would coach students encountering your subject matter for the first time, and then apply those principles to your own learning about AI.
  • You will have to learn to adapt quickly. Artificial intelligence tools are evolving rapidly; your course design and instructional approach will do so, too. Each semester will require additional revisions to your syllabi to accommodate our increasing use of AI tools and AI’s increasing capacities. Starting to build those muscles now with lower-effort activities will pay off in the long run.
  • You actually know how to do this. Researching? Developing hypotheses? Evaluating resources? Check, check, and check. Iterating, revising, and adapting as you go along? Teaching students how to evaluate resources? Guiding students to think about the definitions of “artificial,” “intelligence,” and “human”? Check all that, too. The skills required to become AI-literate from a teaching and learning perspective are skills you already have. It is just a matter of applying them to this particular challenge/opportunity/problem (however you frame it).

Finally, give yourself and your students some grace. This is a huge part of beginning to learn how to teach and learn in an AI world; most likely, neither you nor your students will be proficient AI practitioners this semester. You may miss an academic integrity issue or overlook good opportunities to use AI in a classroom activity. Your students may not cite AI correctly or may not cite it at all. They may be far more fluent with AI than you are, or they may be too trusting of AI. Whatever happens, try to remember that you all are new at this and, as new learners, you all may take missteps and make mistakes with the technology.

Caroline Egan
Caroline Egan is a Project Manager in the Center for Teaching Excellence and Innovation, supporting instructional training and development for Hopkins faculty, graduate students, post-doctoral fellows, and staff.

[1] Darby, Flower. (27 June 2023). 4 steps to help you plan for ChatGPT in your classroom. The Chronicle of Higher Education. https://www-chronicle-com.proxy1.library.jhu.edu/article/4-steps-to-help-you-plan-for-chatgpt-in-your-classroom

[2] Bruff, D. (19 July 2023). Assignment makeovers in the AI age: Essay edition. Agile learning: Derek Bruff’s blog on teaching and learning. https://derekbruff.org/?p=4105

Selected Resources

From Hopkins:

Additional resources:

Image Source: Unsplash

Panel Discussion: “Teaching and Learning in the Age of Chatbots and Artificial Intelligence”

On April 4th, the Center for Teaching Excellence and Innovation hosted “Teaching and Learning in the Age of Chatbots and Artificial Intelligence,” a panel discussion on the implications of artificial intelligence in Hopkins classrooms. This discussion, open to attendees from all schools and divisions in Hopkins, yielded insights into the opportunities and limitations of Chatbots, particularly ChatGPT; identified ways to frame its pedagogical uses for students and faculty; and gave guidance for integrating it into classrooms.

The five-person panel consisted of Victoria Harms, DAAD Visiting Assistant Professor, History; Austin Heath, PhD Candidate, Philosophy; Mike Todasco, MFA student, Writing Seminars and former PayPal executive; and Opal Sitzman and Timothy Huang, first-year students taking the Reintroduction to Writing seminar with Alex Lewis, a Post-Doctoral Fellow in the University Writing Program who is using ChatGPT in his courses.

The discussion produced several incisive observations about chatbots and their role in higher education classrooms.

Here is a summary of the main points:

  • Teaching and learning: There was broad consensus that instructors should engage in active inquiry into artificial intelligence (AI) with their students and leverage the tool to help students think critically about evaluating texts, the accuracy of texts, and what a Chatbot’s opportunities and limitations are as a source, creator, and partner in their work.
  • A metacognitive tool: Both instructors and students said one of the best ways to use ChatGPT is as a tool to help students think about their learning and knowledge, from helping to improve writing to assessing the substance of texts.
  • Academic Integrity: Panelists thought that the written work produced by ChatGPT fell below standards for a finished product; it could be inaccurate, incorrect, and overly broad.
  • Academic Integrity and Assessments: One student urged faculty to identify the core issues driving the need for assessment and use those ideas to motivate students to produce original work. This assessment design contrasts with more mechanical and easily-plagiarizable assignments.
  • The students were teaching the faculty: Opal and Tim provided a huge amount of guidance to faculty, including recommended readings, results from their individual research projects, and thoughts on assessment design.

And words of wisdom from some of the panelists:

  • Austin Heath urged attendees to conceptualize ChatGPT as “a tool inquiry vs. a received text or received piece” of truth.
  • Opal Sitzman warned against a “tend[ancy] to overestimate ChatGPT’s current prowess.”
  • Mike Todasco compared ChatGPT’s current capabilities to “mansplaining,” with all of attendant drawbacks of the term.

Tim and Opal kicked off the conversation, describing the ways that students are using AI technology. Opal assured people that AI is not a “nefarious actor” in student lives: “In general, students like playing around with it like writing a Seinfeld episode, but it’s used more for inspiration than cheating.” Tim said, “You can use it to create the first draft of a paper,” and he’s using it as a self-tutoring tool “to adjust how I write.” Mike, in his MFA classes, used it “to be the voice of a computer in a story I was writing. The key is to always acknowledge it.”

Austin and Victoria discussed how they are guiding students to use and think about artificial intelligence. Austin thought of Chatbots “as a student’s student,” a way for students to learn how to evaluate and critique writing. He gives students output from a chatbot explaining a concept and invites them to grade it and offer suggestions for improvement. In Victoria’s class on Europe since 1945, she asked the Chatbot, “Why did the Soviet Union collapse?” Her students critique the answer for “accuracy and substance,” which taught “students that they know something, too.” She urged the audience “to teach students to be critical digesters of information.”

The panelists also weighed in on how their subject matter expertise influenced the way they used and thought about artificial intelligence. Mike, who has been writing about it for a while, said, “I felt like a Cassandra in that no one was listening and now everyone is talking about it.” He then talked about how “People who don’t have access to JHU resources can use it to learn […] the more people use it – not just for teaching, but for life – will help us learn.” Victoria teaches her students “to fact check results, like I do with Wikipedia. We need to integrate these tools into our assessments so they will use them appropriately.”

Opal, who’s interested in neuroscience, wrote a paper considering whether AI is conscious. Her verdict: “[I]t’s still much more simple than our brain,” but, importantly, “it helps us understand the concept of consciousness even if it isn’t conscious itself.” Austin, as a philosopher, applauded Opal’s interest in consciousness before explaining his own interest in “generat[ing] alternative thoughts about writing and giving credit,” saying, “I’m interested in exploring what it means to give attribution. Did a student write this work? Or did AI write this? Or did students work with AI to write this?”

When queried about Chatbots and academic integrity, the panelists mostly talked about its limitations as an easily accessible cheating tool. Opal said, “ChatGPT has a bad reputation for helping students cheat, but people overestimate its abilities. You still have to do a lot of work that requires critical thinking when using it because it doesn’t produce sophisticated results. It might help with a basic prompt.” Mike and Victoria echoed Opal’s opinion. Mike said, “If you were teaching middle schoolers, you might be concerned with cheating,” though he went on to add, “That said, the future version will get better.” Victoria added, “The pandemic taught us that not all students are excited about technology or are tech savvy.”

Tim offered a very good, thoughtful response about using ChatGPT to plagiarize code in a computing course when Kwame Kutton, a Lecturer in Biomedical Engineering, raised a question about doing this. Currently in a computer science course himself, Tim said, “In BME there are unique opportunities to write code that saves lives. Therefore, students need to tackle the core issue to solve before they even write code. We want faculty to teach us how to think about the logic of the problem, not just writing code.” His comment encouraged instructors to think deeply about first framing and identifying the problem for students, which will help motivate them to produce original and independent work.

Mike stated another perspective: “I don’t know any programmer who doesn’t use Copilot,” a code repository on GitHub that uses AI to suggest coding solutions. “My analogy is calculators,” he said. “You need to know how to do math without a calculator, but once you are doing the calculations after setting up the problem, you should use a calculator to help solve the problem.”

A question from the audience about languages, accents, and ChatGPT turned the discussion to issues of accessibility and political bias. Tim saw one of his friends using the Chatbot to translate English to Japanese and then used it himself to translate a Spanish article he was familiar with. His opinion: “It does a better job than Google Translate” though “there are lots of metaphors that get lost in translation by these tools.”

Mike then gave two excellent examples about how ChatGPT is providing access and support to people with divergent and impaired abilities. He said, “ChatGPT 4 is available, but they haven’t released the picture-to-text feature that exists yet. They shared video of someone with visual impairment using ChatGPT 4 to learn what was in the fridge using their phone. It will be able to do amazing things in the future to help us.” He went on to talk about a friend of his who knew someone in San Francisco with a lawncare business who struggled to communicate via email. The owner of the business now uses ChatGPT “to help polish his emails,” thus improving his client relationships.

Opal talked about how ChatGPT struggles with dialects, which turned the conversation to political bias. She’s using ChatGPT to write a short story “in the style of Kate Chopin,” a 19th Century American writer known for writing about Louisiana Creole culture. Opal said, “[Chopin] used a lot of Louisiana dialect” and ChatGPT “struggles” with this because it “is filtered so it doesn’t mimic the racist language used during that time.” She said that people have found ChatGPT to be “an establishment liberal” in its political biases. Victoria brought up “issues of bias in Silicon Valley” and wondered how ChatGPT would address Critical Race Theory (CRT). Mike decided to ask ChatGPT whether we should ban CRT and copied and pasted ChatGPT’s response in the Zoom chat:

As an AI language model, I don’t have personal opinions. However, I can provide you with an analysis. Whether Critical Race Theory (CRT) should be banned in schools is a subjective question, often debated among educators, policymakers, and parents. Supporters argue that CRT promotes understanding of systemic racism and its impacts, while opponents believe it can be divisive and foster racial animosity. Ultimately, the decision to include or exclude CRT in schools depends on the goals and values of the educational community involved.[1]

The conversation ended with speculation about how quickly ChatGPT would progress. Mike said, “The current GPT4 has been remarkable. I’ve written fiction in each version and I’d say it’s getting two grade levels better in each version.” Opal also weighed in: “It will be quick, but I’m not wary yet. We need to keep considering these questions, but I think it’s less something to be scared of and more something to utilize. I don’t see anything being more powerful than humans in the near future.”

Recommended reading and activities:

[1] OpenAI. (2023). ChatGPT (Apr 4 version) [Large language model]. https://chat.openai.com/

Caroline Egan
Caroline Egan is a Project Manager in the Center for Teaching Excellence and Innovation, supporting instructional training and development for Hopkins faculty, graduate students, post-doctoral fellows, and staff.

Image source: Pixabay, Unsplash

ChatGPT: A Brief Introduction and Considerations for Academic Integrity

I’ve been reading about the potential impact of artificial intelligence (AI) on teaching and learning for some time. A close family friend gave me a book entitled In the Mind of the Machine by Ken Warwick in 1998. The Education Horizon Report Advisory Committee, of which I was a member, first listed artificial intelligence as an emerging technology likely to have an impact on learning, teaching, and creative inquiry in education in 2017. November 2022 brought the long-anticipated arrival of ChatGPT beta with accompanying media attention.

What is OpenAI and ChatGPT?

OpenAI is an artificial intelligence research lab. Open AI developed a chatbot called ChatGPT (GPT = Generative Pre-trained Transformer) and an image generation application DALL-E 2. ChatGPT is a natural language processing model that is trained on hundreds of billions of documents and websites. Users can interact by asking it questions or submitting statements to which it can generate responses.  For example, here is ChatGPT answering a question about itself:

Prompt: What is ChatGPT?
“ChatGPT is a large language model developed by OpenAI. It is trained on a diverse range of internet text and is able to generate human-like text in response to various prompts. The model can be fine-tuned for various natural language processing tasks such as language translation, question answering, and conversation.”

While ChatGPT received most of the media attention in winter 2022-23, there are other chatbots that exist like Jasper and Chincilla.

What are the main concerns?

The main concern for instructors is students asking OpenAI applications to complete assignments for them. This includes writing essays or research papers along with coding assignments for which ChatGPT is trained. Students can also ask ChatGPT to answer test questions.

Things to Consider

While the capabilities of artificial intelligence applications will continue to evolve, there are currently some limitations. For example, current models do not include articles behind paywalls (e.g., subscription journals). This makes it harder for students to generate essays based on peer-reviewed research.  While the models are trained on a large number of documents, the applications’ responses to specific, focused inquiries tend to be vague.  My colleagues and I asked ChatGPT to write a strategic plan for the Center for Teaching Excellence and Innovation. It suggested relevant ideas, but it was generic and too broad to be useful. That said, we could have used it as a starting point for brainstorming a draft.

Some applications, like Turnitin, are claiming they can detect if students used ChatGPT, but like any technology, these applications are not perfect and students can work around them (e.g., editing the essay produced to make it closer to their own writing style).

 Academic Integrity

Use of OpenAI applications can fall under academic integrity policies like plagiarism, but the gray zone between clearly plagiarized work and an academic support tool is large. For example, most instructors would consider it plagiarism for students to ask ChatGPT to write a paper based on a writing prompt from class. But is it OK for students to ask ChatGPT for a summary of research on a topic, which they then use to generate a bibliography as the basis for a research paper they write?  Instructors should learn more about how ChatGPT and other AI technologies work so they can inform students what is considered appropriate use of AI technologies and what is not. Here are additional strategies to consider to help you and your students navigate this new territory:

  • Scaffold the activity by asking students to turn in an outline and iterative drafts that address comments and feedback from the instructor or teaching assistants. This requires students to show progression in a way that is difficult for tools like ChatGPT to produce.
  • Ask students to write papers using a shared Microsoft document through One Drive so you can see the version history.
  • Use writing prompts that are more specific or require students to cite specific texts.
  • Use AI tools to teach students. For example, generate essays in ChatGPT and have students critique them.
  • Discuss with students what is considered acceptable use of AI technologies (e.g., generating a summary of a field) and what is not (e.g., responding to a specific assignment prompt).

A colleague also commented that as we engage with ChatGPT and other AI technologies, we are feeding it data it can use to improve its models. They own the submissions as part of the terms of agreement when accounts are created. Explain to students that they may be giving over their intellectual property if they are using these tools.  If they submit your tests for ChatGPT to answer, they may be violating your intellectual property rights.

Where to Learn More

 Here are some resources to learn more about AI technologies:

We are all orienting ourselves to this new technology and best practices are evolving. The CTEI will continue to share more information and host discussions over the semester.

Mike Reese
Mike Reese is Associate Dean of the Center for Teaching Excellence and Innovation and associate teaching professor in Sociology.

Image Source: OpenAI Logo, Pixabay