Thursday, February 8, 2024
8–9 am EDT
Recommended Guidelines and Ethical Considerations for Teaching and Learning with Generative AI
This interactive conversation will discuss recommended guidelines and ethical considerations essential for both instructors and students engaged in the teaching and learning of generative AI. The presentation will cover four key areas:
- Instructor Guidelines: Recommendations for educators to consider.
- Student Guidelines: Recommendations tailored specifically for students learning about generative AI.
- Generative AI Ethics & Pitfalls: An outline of ethical dos and don’ts, emphasizing the importance of ethical decision-making in the use of generative AI.
De Angela L. Duff – Associate Vice Provost, NYU, and Industry Professor, NYU Tandon
9–10am EDT
Student-Centered Learning Through the Lens of Generative AI
Student-Centered Learning (SCL) tailors instruction to individual interests and abilities, promoting deeper conceptual learning and higher-order skills like collaboration and critical thinking. The use of Generative AI presents opportunities and challenges for SCL, requiring instructors to rethink active learning strategies, consider the learning environment’s impact on students’ cognitive effort and motivation, and facilitate metacognition, scaffolding, and assessment. This interactive conversation explores SCL dimensions through Generative AI, helping instructors enhance their own SCL practices.
Maaike Bouwmeester – Clinical Assistant Professor, Educational Communication and Technology, NYU Steinhardt
10am–11:30am EDT
NYU Faculty Lightning Presentations Round #1
The AI Disclosure Statement
Believing that we do not currently have the ability to police or prohibit students’ use of AI, in 2023 I piloted a course policy that I call the “AI Disclosure Statement”: students may use AI however they want, provided they submit a statement detailing its use and any lessons learned. With this talk, I will outline the basis for and substance of this policy, findings from one semester’s implementation, and the preconditions I endorse for successfully employing an AI policy aimed at fostering trust and learning– both academic and practical.
Celia Wright – Adjunct Faculty, School of Global Public Health
Avoiding Generative AI in Students’ Graded Writing: Ten Assignment Types That Don’t Work and Three That Do
For those who would like to create assignments that avoid the temptation for students to use generative AI fraudulently, I will illustrate some of the types of graded writing that instructors should steer clear of. The ten types of take-home work in question focus on examples of purportedly “AI-proof” assignments that are, in reality, wholly or largely within the capabilities of current AI programs. I will conclude with a discussion of three assignment types that I have found to be robustly resistant to cheating with generative AI.
Matthew Stokes – Adjunct Faculty, Liberal Studies, NYU Arts & Science
Using Generative AI for Explaining AI Decisions
Drawing on my experience with Large Language Models (LLMs) and AI in autonomous robot safety, there’s a unique opportunity to shape how AI tools are used in education. We can set clear guidelines for students on AI’s acceptable use, focusing on critical thinking and originality. Educators can use generative AI to create innovative assignment prompts and tailor assessment scenarios for individuals or teams to improve efficiency. This approach enhances learning but still challenges identifying AI-generated versus human-generated content. It’s about leveraging AI to innovate in education while maintaining academic integrity and deepening students’ AI understanding.
Aliasghar Arab – Adjunct Faculty, Mechanical and Aerospace Engineering, NYU Tandon, Director of Agile Safe Autonomous Systems (ASAS) labs.
The ABCs of Evaluating GPTs: Evaluating Generative AI Tools for Academic Research
In our role as Academic Librarians at NYUAD Library, we teach sessions within courses that cover various information literacy skills, such as finding and evaluating sources, building citations and ‘citation chaining,’, finding highly cited papers, avoiding plagiarism and supporting faculty and researchers with altmetrics and scholarly metrics. When generative AI came on the scene, we had to find a way to integrate it into our sessions, especially after learning many students had embraced and were actively using these tools, sometimes inappropriately or unethically. Faculty differ in their approaches to AI use in the classroom. As Academic Librarians, we evaluate these tools through the lens of information literacy and make recommendations based on their appropriateness for different stages of the research process. We’ll explore three different Generative AI tools that we have vetted for their applicability in an academic research setting and how we evaluated them.
Amani Magid – Coordinator of Instruction and Associate Academic Librarian for the Sciences and Engineering, NYUAD Library
Grace Adeneye – Librarian for Arts and Instruction, NYUAD Library
Developing AI Assistants for Higher Ed Instructors – Examples
I am currently developing GenAI tools for instructors who are teaching business courses. There are 3 applications: (1) An auto-grading tool that can grade open-response questions in exams based on the instructors’ grading criteria and then offer an analytic summary of the student responses. Existing products in the market do not offer such features. The ultimate goal is to build this app on the internal servers to protect the student data. (2) A verbal feedback tool that can summarize the instructors’ verbal feedback and then track the students’ progress. Our observation is that professors generally have limited time to give comprehensive written feedback for students working on Capstone projects. A lot of information is lost in verbal feedback. So having an AI summary tool can mitigate the expectation gaps (3) A RolePlay tool for business courses like Negotiation and Human Resource Management. Business cases provide a wide range of context and knowledge for scenario simulations in which students can have an immersive learning experience. All three tools are still under development.
Nicole Wang – Assistant Professor of Practice, Interactive Media and Business, NYU Shanghai
Making ShiffBot
ShiffBot is an experimental AI chatbot built in collaboration with Google Creative Lab using Google’s Gemini large language model augmented with an educational corpus drawn from my own course materials to help learners when they’re working on their own in p5.js. I’ll demonstrate how we built on top of the Gemini API using prompt engineering for adjusting the teaching style and embeddings and semantic retrieval to pull from a reliable and relevant knowledge-base.
Dan Shiffman – Associate Arts Professor, Interactive Telecommunications Program (ITP), NYU Tisch
Moderated by Lucy Appert – Director, Office of Educational Technology, NYU Arts & Science
11:30–12:30pm EDT Break for Lunch
12:30–1:30pm EDT
NYU Student Roundtable
This session was NOT recorded!
Moderated by Hui Soo Chae – Executive Director, Learning and Teaching Nexus, and Clinical Professor, Education and Technology, NYU SPS
Maan Algarzae – B.F.A. Interactive Media Arts (IMA),
NYU Tisch ’27
Priyanka Bose – M.S. Computer Engineering, NYU Tandon ‘24
Iti Saanchie Goswamy – M.S. Integrated Design & Media, NYU Tandon ’25
Phillip Kantorovich – B.S. Business & Technology Management, NYU Tandon ’26
1:30–3pm EDT
NYU Faculty Lightning Presentations Round #2
De-Automating Writing: Digital Tools for Assessing the Likelihood of Student Cheating
My talk describes digital tools I’ve developed to help faculty assess the likelihood of student cheating on given assignments or in given courses. “ChadBot,” a custom GPT developed using OpenAI’s new platform, is instructed to adopt the voice and manner of a college freshman, who “totally gets why students cheat,” and yet has deep familiarity with decades of research on cheating and learning. That’s because this GPT is trained on scholarly research from cognitive psychology, education, and writing studies.
When provided a college-level a college-level writing prompt, this GPT draws on this research in order to characterize the overall likelihood of student cheating, and then provides three short-term, three medium-term, and three long-term plans for revising the writing prompt (or the classroom environment) in order to better promote student engagement, and make cheating appear less necessary and less attractive to students.
Alexander Landfair – Clinical Associate Professor, Expository Writing Program, Arts and Science
Generative AI to Support Real-World Projects
In this talk, Drs. Steven Goss and Paul Acquaro explore the potential of generative AI to support real-world project implementation, focusing on its role as a substitute for stakeholders providing feedback on applied projects. They will discuss the development of two specialized bots: a Subject Knowledge Bot (SKB) and a Project Knowledge Bot (PKB). The SKB, aligned with course theories and strategies, offers content-knowledge feedback to students as they work towards the project goals. Complimenting the SKB, the PKB, trained with knowledge from the stakeholders, guides the students on the project’s stated requirements. These bots will be used as part of an applied project assignment where students will work to redesign an internal communication document. The faculty plans to use the findings from this activity to understand how generative AI bots can scaffold real-world project implementation.
Steven Goss – Chair and Clinical Assistant Professor of the Management and Technology, Division of Programs in Business-Management and Technology, NYU School of Professional Studies
Paul Acquaro – Adjunct Assistant Professor, Division of Applied Undergraduate Studies, NYU School of Professional Studies
Understanding the Team Dynamics through a Virtual Project Manager: A Case Study of Generative AI
In engineering education, fostering collaborative skills among students is crucial, and team-based learning has become the primary approach. It is an approach particularly prevalent in foundational courses, such as first-year cornerstone courses and senior-year capstone design courses, but it also finds application across the entire engineering curriculum. The overarching goal for implementing team-based learning is to enhance student’s abilities to work effectively in groups, aligning both with the demands of their future professional endeavors and broader educational objectives. In a large engineering course, instructors interact with more than 100 students daily. A virtual project manager would not only share the burden with the faculty members, but identify potential team issues promptly, provide constructive/personalized feedback to student teams, and predict milestone team performance.
Rui Li – Industry Assistant Professor, General Engineering/Mechanical Engineering, NYU Tandon
Simple Ways to Teach How Generative AIs Work
I will go over the Peanut Butter And… approach (a metaphor for how LLMs function), Contexto.me (a game that can teach how to think like an AI), and Uncannny Valley (the card game to welcome our AI overlords)
Barry Joseph – Adjunct Faculty, Learning Technology and Experience Design, NYU Steinhardt
Exploring the Evolution of AI Technology in Video Production
In this session, the presenters will showcase advancements in AI video generation technologies that have rapidly evolved. They will discuss educational uses of AI-video generation tools for assignment feedback, role-playing, and fast translations. The presenters will also review the tools and platforms they used to develop these educational resources and highlight the limitations and potential of these tools.
Kristine Rodriguez Kerr – Academic Director, MS in Professional Writing, and Clinical Associate Professor, Professional Writing, NYU School of Professional Studies (SPS)
Hui Soo Chae – Executive Director, Learning and Teaching Nexus, and Clinical Professor, Education and Technology, NYU SPS
Negar Farakish – Assistant Dean, Division of Programs in Business, and Clinical Associate Professor, NYU SPS
B-Roll: Open Source AI Short Filmmaking
B-Roll is a workshop focused on the practical use of open-source AI tools in filmmaking. In this session, we’ll be utilizing tools like Stable Diffusion, ComfyUI, and Bark to create super short works of media art.
I would like to teach participants a potential workflow for working with generative AI art tools for film, photography, and sound. My goal is to run this all off my laptop and for participants to have a nice step-by-step way of recreating my workflow.
Christian Grewell – Clinical Associate Professor of Technology, Operations, and Statistics, NYU Stern
Moderated by Anandi Nagarajan – Assistant Vice Provost for Pedagogy, NYU Office of the Provost
3–4pm EDT
Educator-to-Educator Conversations:
Teaching & Learning with Generative AI
This session was NOT recorded!
Engage actively with your peers in Zoom breakout rooms to explore generative AI’s impact on teaching and learning. This is an invitation to dive into meaningful discussions about the nuances of integrating generative AI into teaching and learning practices. Your active involvement is crucial for maximizing the benefits of this knowledge exchange.
De Angela L. Duff – Associate Vice Provost, NYU, and Industry Professor, NYU Tandon
NYU Stream Video Playlist