Pioneering AI Use In The Business Analytics Classroom by: Gies College of Business on November 01, 2023 | 440 Views November 1, 2023 Copy Link Share on Facebook Share on Twitter Email Share on LinkedIn Share on WhatsApp Share on Reddit Vishal Sachdev has been teaching IT for Networked Organizations at Gies College of Business for years. But this year, he chose to embark on a bold experiment: harnessing the power of artificial intelligence and large language models like ChatGPT to transform the way students learn. The course examines information technology and its impact on modern organizations – essentially exploring the business value of technology, with an exposure to emerging tech. This includes ChatGPT, an AI chatbot that can pen essays, write songs and even generate computer code from a few simple prompts. Its advanced capabilities are leading some instructors to ban its use in the classroom, but for Sachdev, it’s much like earlier efforts to stop computers and calculators from entering the door. In the end, those technologies became tools that furthered education. So why not embrace change and make that leap now? “This is pretty much going to disrupt everything we do in education,” said Sachdev, who is part of a faculty working group at Gies dedicated to exploring the use of AI in the classroom. “So, the only way to figure it out is to experiment.” And that’s what he did, securing a grant to provision server space, hire developers, and implement the components necessary to make AI part of the learning experience. “The initial idea was to look for an interface where I could have multiple people talking to each other and have the bot as support, kind of like a teammate, just with infinitely better memory,” said Sachdev. However, that wasn’t possible, because most bots are designed for a single-user experience. So he went for the next best thing, using a bot named Erin that could interface with students on Discord, a communication platform which provides a multi-user chat experience. Using that combination, he could create private communication channels where each student could interact with a bot while allowing him to observe those interactions. He could also provide public channels where multiple students could interact with each other and the bot. Sachdev and fellow Gies professor Unnati Narang recently received a $7,500 grant from the Provost’s Initiative on Teaching Advancement to explore whether different types of interactions with the bot (individual vs. group) could lead to different learning outcomes in Gies’ Master of Science in Business Analytics (MSBA) program. He teaches a database management course and serves as academic director of the nine-month, STEM-designated MSBA. The program prepares students through a strong integration of business application area expertise such as finance, information systems, and accountancy, and data science skills including gathering, managing, analyzing, and communicating data. Graduates leverage business analytics methods and tools to solve complex business problems so they’re prepare to succeed in their careers from day one. Sachdev’s AI experiments started with a Q&A bot based on the course syllabus that allowed students to become familiar with the basic outline of the course, asking questions about grading and deadlines and getting answers in real-time. “Each student is essentially leveraging the knowledge base of the bot, without any custom training from me,” said Sachdev. “But we also have other channels, which are set up for a particular activity where we did train the bot, and that was made available to all students to work together.” This includes a bot that designed to act as a debating partner. For this experiment, Sachdev developed two different channels, one taking the pro side of an argument and the other taking the con. From there, students could interact with the bot, honing their debating skills and fine- tuning their positions. In another experiment, Sachdev developed a bot that could critique student submissions, training it with assignment instructions, a rubric, a model submission, and guidelines on how to respond to students. “It can offer an evaluation on each point in the rubric and give the students guidelines, like ‘hey, you might be missing something,’ or ‘your word length might be off,’ or ‘you haven’t really answered the question,’” said Sachdev. Using this tool students can then improve their work through multiple iterations. Sachdev is currently building a more robust prototype of this tool designed for faculty in high- enrollment courses. In recognition of those efforts, he was recently named one of eight Teaching and Learning Faculty Scholars for the 2023-2024 academy year by the Gies College of Business. Experiments like these could one day lead to a virtual TA who’s fully versed on the course, never tires of answering questions, and is available to students 24-hours a day. On the teaching side, AI can also help instructors in large classes create personalized learning activities. As proof of this concept, Sachdev created an activity where the AI bot could coach a student on SQL, a key skill for analytics, and customize the level and context for each student in a class of 50. The student initiates the conversion about learning SQL, and the AI bot responds with a prompt for the student to specify an industry with which they want to work. If a student picks music, the bot responds with a database for a music website like Spotify, creates some synthetic data, and proceeds to ask the student ten business questions for insights from the data. The student provides the SQL queries, and the bot simulates the execution of the query on the sample data and provides feedback. He did that for 50 students, but with AI, that could easily be scaled up to hundreds of students, helping instructors create custom practice questions while also enhancing engagement. Eventually, it is possible that AI could go well beyond that, creating a custom learning path for each student. Sachdev said advances in AI are moving so quickly that they won’t just change the classroom, they’ll transform teaching itself. “It’s now trivial to train a bot with a 700-page chemistry book,” said Sachdev. “Now stretch that out and assume that all knowledge is instantaneously available to everyone, and soon compressed enough that it can fit on your phone as well. Then what is the role of a teacher?” “We have known that we are no longer the sage on the stage, but a guide on the side,” said Sachdev. “But I think we’re getting a little further out, where you’re essentially a curator of an experience that you want the students to have. As the knowledge base becomes better, and your ability to curate that experience becomes easier, then all you’re really worried about is making sure that students are meeting the learning objectives and supporting that journey.” According to Sachdev, it’s only natural that ground-breaking experiments like these are happening at Gies. “We’ve always been the first to try new things,” said Sachdev, citing a long list of accomplishments as proof, from creating the world’s most affordable and accessible online MBA program, to a Disruption Lab that’s pushing the boundaries of virtual and augmented reality. “I mean, I’m the director of a 3D printing lab, right? Why is that set up in a business school? Because we wanted to see whether ideation from an idea to a product could happen faster. And business students are responsible for ideas. So, why not here?”