Photo by Aubreonna Chamberlain/DePaul University

​The AI in Teaching Symposiums are opportunities for DePaul faculty and staff to share and see how their colleagues have successfully utilized an AI activity or assignment into their teaching. Quick, simple, direct demonstrations that can be easily adopted in other courses at the university.

​The second AI in Teaching Symposium took place Friday October 18th, 2024 at 1PM.

• 1:00 PM: Opening Remarks.
• 1:15 PM: Stefanie Demetriades
• I Think, Therefore AI: An experimental assignment writing about AI, with AI
• 1:30 PM: An Chih Cheng
• AI Powered Exam Prep (Adaptive Testing System)
• 1:45 PM: Amy Merrick
• AI as a Bias Detector
• 2:00 PM: Adam Gao
• Building Your Own Customized GPT Teaching Assistant for "Free"
• 2:15 PM: Q&A

Handouts can be downloaded here.

NotebookLM was used to create a briefing document from the day (from the transcript and handouts). This has been lightly edited. Additionally, here is a link to an NotebookLM-created podcast, based around the symposium presentations and Q&A. This was created using the transcript and PDF handout.

Main Themes:

• AI as a double-edged sword in education: The symposium explored the potential of AI tools like ChatGPT to enhance learning while acknowledging the challenges they pose, particularly regarding academic integrity.
• AI for personalized learning and assessment: Presenters demonstrated how AI could be used to create customized learning experiences, generate high-quality assessment materials, and simulate real-world testing scenarios.
• Prompt engineering and critical thinking: The importance of prompt engineering and developing students' critical thinking skills when using AI was emphasized throughout the symposium.
• Ethical and societal implications of AI: The discussion extended beyond the classroom, highlighting the need to consider the environmental impact and potential biases of AI systems.

Key Takeaways:

Stephanie's Presentation:

• Experimental assignment "I Think, Therefore, AI": Stephanie detailed an assignment requiring students to write an op-ed using AI, prompting critical reflection on its use.
• Students identified AI's strengths in structuring and brainstorming, but noted its lack of personalization and tendency to "hallucinate" evidence.
• Key Takeaways for educators: Encourage nuanced and critical engagement with AI tools.
• Don't demonize AI, but guide students towards thoughtful and informed decisions about its use.
• Design assignments that leverage AI's strengths while emphasizing student ownership and critical thinking.

Quotes:

• "Encouraging ownership of your work was a key theme that I really wanted to emphasize, that this is a tool that you can figure out and choose to use in various ways. But at the end of the day it is your work right?"
• "Students found real value in the kind of structuring brainstorming and organizational aspects of AI, helping them to get their thoughts in order to try out a few different structures."
• "I don't think it's comparable to what I would have produced without AI. So recognizing, starting to pull apart this, yes, AI, no AI, and pulling apart into a more nuanced practical application of Where does it help? Where does it hurt?"

An Chih's Presentation:

• Leveraging AI for content mastery and licensure exam preparation: An Chi described a two-step approach using AI to create customized quizzes and simulate adaptive testing environments.
• By combining Bloom's Taxonomy with careful prompt engineering, high-quality question banks can be generated.
• AI-powered adaptive algorithms personalize the testing experience, focusing on areas where students need further development.

Quotes:

• "The problem with these, even though, even if they are high quality quizzes, you can easily find them online like quizzes. And now they GPT everything. You don't need even need to type. You just copy the screen, and you paste to GPT to give you the answer."
• "So first, you would need some kind of framework like Bloom's taxonomy that specific guide the GPT through proper prompt engineering."
• "So you can see, student does not need to take a thousand questions sitting there for 5 h instead of just sitting there 1 h, maybe taking only 30 questions. The system automatically zoom in to find the level, that where student is."

Amy's Presentation:

• Exploring AI for identifying bias in writing: Amy presented a collaborative exercise where students and AI worked together to identify and address potential biases in written work.
• While both students and AI offered valuable suggestions, the AI often took a more holistic approach, evaluating the overall narrative.
• The experiment highlighted the potential of AI as a tool for promoting inclusivity and strengthening writing, although ethical concerns remained.

Quotes:

• "Could AI be helpful in identifying biases in our writing?"
• "On the right side of the screen. There were a few differences, so students were more likely to go word by word and look at adjectives that maybe were emotionally laden, or that introduced opinion and kind of look at those changes, whereas the AI tended to more evaluate the story as a whole."
• "AI leaves a bad taste in my mouth, though it will be some time before I trust it enough, and feel morally okay enough to use it in this manner, even if the suggestions are excellent."

Adam's Presentation:

• Introducing "Vita," a custom GPT teaching assistant: Adam showcased "Vita," a customized GPT model designed to answer student questions about course content, logistics, and programming assignments.
• While acknowledging the potential for misuse, Adam emphasized the importance of setting boundaries and clear expectations for both students and the AI.

Quotes:

• "Building Your Own Customized GPT Teaching Assistant for 'Free'"
• "It's role is to try to help students with their understanding of the course materials and the programming assignments, how to debug and answer some course, logistic questions."
• "You cannot just chat with Vita for any topics."

General Discussion

• Pervasiveness of AI-assisted cheating and evolving definitions of academic integrity.
• The need for institutional policies and resources to address AI's impact on teaching and learning.
• The environmental cost of AI and its potential for exacerbating existing social inequalities.

Further Research and Action Items:

• Explore best practices for prompt engineering and integrating AI tools into diverse learning environments.
• Develop assessment strategies that account for the availability of AI assistance while encouraging critical thinking and originality.
• Engage in open dialogue about the ethical implications of AI in education and its impact on society as a whole.
• Investigate the feasibility of university-wide subscriptions to AI tools and resources for faculty and students.
• Continue research into more energy-efficient AI models and infrastructure to mitigate the environmental impact.

This briefing document highlights the key themes and takeaways from the AI and Teaching Symposium, offering a starting point for further discussion and action. As AI continues to evolve, it is crucial for educators to engage critically with these tools, collaborating to maximize their potential while mitigating the risks.