• Skip to main content
  • Skip to footer

Center for Teaching & Learning | Baruch College

  • About Us
    • Our Team
    • Appointments
    • Faculty Advisory Committee
    • Contact Us
  • Events
  • Resources
    • Pedagogy
    • Technology
    • Research
    • AI Resources
  • Appointments
You are here: Home / AI-Resources / Incorporating AI in Project Planning at the Undergraduate and Graduate Level

Incorporating AI in Project Planning at the Undergraduate and Graduate Level

Filed Under: AI-Resources June 26, 2025 by Editorial Team

Liz Minei, Department of Communication Studies

Tags: AI for brainstorming, project development, thesis and capstone planning

Introduction

Courses that require the execution of project planning for students spanning the semester often require the instructor to teach students to engage with the “devil’s advocate” as a means of challenging cognitive bias and avoiding groupthink, especially since the tendency to fall into groupthink results in poor outcomes and decisions that are made too quickly in favor of keeping the peace. Sometimes, challenging cognitive bias requires assigning other students the role of critical thinker of a proposed plan, allowing for review and feedback in dyadic settings. Other times, such as group or team situations, this method is carried out as a portion of the planning stages of the project assigned by the professor. Many times, the professor is placed in the role of offering critical feedback or suggestions along the lines of “what if X occurs, what is the backup plan?” This AI artifact can give professors a chance to guide students through the process of autonomy in the long-term planning and execution of a guided project (e.g., group or teamwork at the undergraduate level or thesis and capstone students at the graduate program), particularly because generative AI, like ChatGPT can give openings to challenge biases if asked to do so. In other words, students can be coached to think of ChatGPT as a potential well-meaning critic of their work, who might just open their eyes to flaws that can be addressed in advance. 

Downloads

Sample Instructions for Generating Topics for a Final Master’s Thesis or Capstone (PDF)

faculty photo

Professor Minei is an Associate Professor of Organizational and Interpersonal Communication. Her research interests include leadership, High Reliability Teams, group communication, and marginalized identity in the workplace. Her teaching style involves a hands-on learning approach, experiential learning, and game-play for application. 

Sample Class 1: Group Communication (Undergraduate) Com 3078

Students are required to apply group communication principles to a philanthropy project spanning the last 8 weeks of class. Students learn how to combat groupthink by challenging various parts of their projects in the planning stages, and are required to answer 3 cognitive bias questions (covered in class; e.g., “Tell me where this plan can go wrong/Strengths and Weaknesses of this plan?” or “What are potential (external) disruptions and how can we plan for/avoid that disruption NOW?”or “How might someone outside our group (mis)perceive our actions?” or “What ideas might be outside the realm of possibility? What if funds were unlimited?”) as a means of challenging groupthink and as a best-practice for completing the project. 

Uses of AI

Students can be guided to input portions of their project plan into a generative AI program like ChatGPT and ask specific questions that analyze the plan as they have created it. Example questions can be:

  1. Review this plan for feasibility-what portions can be completed in 6 weeks and what portions are beyond the scope of the time frame. 
  2. Review this plan from an impression management standpoint. How might this plan create tensions for recipients that might not be apparent at this time (ex: students that want to partner with members of the homeless population are often unaware that their plans to film homeless individuals being gifted some element can be considered insensitive).
  3. What flaws, considerations, or concerns did AI uncover about your groups’ decision that prompted you to alter some element of your plan, or create a contingency?
Notes For Faculty

The AI portion of this assignment was created and added on to two assignments, in different sections, that had already been run for years without the AI portion. The intention to add these AI prompts, particularly in courses where students would be engaging in brainstorming, was to get them to think through their project from the perspective of challenging their own ideas. Groups, especially undergraduates in the planning phase of a project, tend to reach consensus very quickly as a means of keeping the peace. Asking them to run their “decision” through AI-guided questions can bring awareness to the ways in which their initial ideas might be good, but flawed.  


Sample Class 2: Final Project Bootcamp(Graduate Level) Com 9960

Students in the Masters Final Project Bootcamp class spend the semester creating the first portion of their project outline from brainstorming ideas about potential topics, all the way to the creation of a start-to-finish project outline with portions of written text (e.g., rationale and introduction). Often, the instructor serves in an advisor capacity, asking questions like “What are the goals of the project?” and “How are you planning on managing data to address the scope of the research questions?” 

Uses of AI

Students can extend their ideas further by using generative AI like ChatGPT to assist in challenging their ideas, rather than generating the ideas. Students can enter portions of their assignments into the prompt and ask questions that can assist in challenging or building on ideas for guidance. Examples of questions can be “what other peer-reviewed readings can assist me in developing this topic further? Please search the internet for articles and provide the links to the articles.” 

Sample Flow

The following screenshots demonstrate the kind of prompt students can answer as well as the dialogue presented by ChatGPT. 

Photo 1: The first screenshot below shows the kind of prompt that students might use to start the conversation. Instructors can encourage students in any number of directions concerning what kinds of information to load into the ChatGPT question field. Below, the prompt is fairly simple but students can add additional details. 

screenshot of ChatGPT prompt: "We are undergraduate group communication students at Baruch college working on a group project for the end of semester project. there are 4 members of the group and we decided that we want to create a philanthropy project that brings donations to a women's shelter in NYC. We have 6 weeks from start to finish to create this project. Please help us think through potential pitfalls or areas that we might encounter issues."
screenshot 1
screenshot of ChatGPT reply: "That's a meaningful and ambitious project—well done for choosing something with community impact! A six-week timeline is doable, but there are several areas where student groups often encounter challenges. Here's a breakdown of potential pitfalls and how to plan for them:"
screenshot 2
screenshot of ChatGPT output of group work pitfalls, including "#3. Shelter Parternships" and "4. Fundraising/Donation Logics"
screenshot 3
More group work pitfalls, numbers 5 ("Promotion & Audience Engagement") and 6 ("School Guidelines/Approvals")
screenshot 4
Screenshots 2-4: Students can see from the responses of ChatGPT that AI has picked up on the prompt of “pitfalls”. They can follow up with another question like, “Develop a contingency plan for line item 6: School Guidelines and Approvals” or they can ask another question that challenges bias (e.g., What other populations might benefit from large scale donations that might not be considered as vulnerable?” 
pitfall #7, "Final Project & Presentation: Forgetting to document your process or struggling to tell the story."
Screenshot 5: In this photo, we can see that ChatGPT is prompting students with suggestions that push their plans into different data presentations. Students can be encouraged to follow up on some element from their first question, or can be given the freedom to engage with the chat as they need to. 

Footer

Site Navigation

About Us

  • Appointments
  • Contact Us
  • Faculty Advisory Committee
  • Our Team

Events

Featured Projects

  • Active Learning
  • Blogs@Baruch
  • CUNY 1969
  • Student Experience Survey
  • Student Learning Guide
  • Teach Hybrid
  • Teach OER
  • Teach Online
  • Teach Open Tools
  • VOCAT
  • Zoom Guide

Resources

  • Pedagogy
  • Technology
  • Research

Search

Topics

  • Active Learning
  • Anti-Oppressive Pedagogies
  • Assessment
  • Assignment Design
  • Blackboard
  • Blogs@Baruch
  • Classroom Practices
  • COIL
  • Copyright
  • Course Design
  • Discussion
  • Ed-Tech
  • Experiential Learning
  • Faculty
  • Game-based Learning
  • Group Work
  • Hybrid
  • Jumbo Courses
  • Media Literacy
  • OER
  • Online
  • Research
  • SoTL
  • Students
  • VOCAT
  • Zoom

Baruch College Center for Teaching and Learning · 151 E. 25th Street, Room 648 · 646-312-1565 · [email protected]


Creative Commons License
Unless otherwise specified, all CTL site content is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.