By Dr. Holly Whistler, Coordinator, Online Programs, Virginia Community College System
LLM.
A.I.
Machine Learning.
Neural Network.
Chatbot.
These are terms that dominate our LinkedIn feeds and work emails. However, I recently received a notification about an A.I. Symposium at University of Virginia on January 23rd that caught my attention: “Advancing Teaching and Learning in the Age of Generative A.I..” I registered and landed in Dr. Rachel Chung’s session; some of her content included the word “excel” and I wondered how that was connected to A.I..
Pressure is mounting for colleges to articulate exactly how A.I. can benefit their students and impact programs so essential to the future of work. A.I. can feel like an invasive species instead of coevolution. A pervasive dilemma remains for academia: how does an instructor teach something so complex that is evolving so fast?
Dr. Rachel Chung says you can train anyone in 10 minutes a day.
Dr. Chung is a Clinical Professor of Operations and Information Systems in William and Mary’s School of Business and has used her “neural simulator” activity since her initial interview with the college. As a clinical professor, she lives between the didactic and experiential worlds and often hears students lament that they are NOT math majors or computer science phenoms.
“I tell my students that (excel) formulas are coding – so you CAN code.”
Excel is the “language of business” and in her world as a business analytics professor, she has to move her students past abstract and into concrete use cases. Her lessons mimic A.I. through spreadsheet based case studies – leveraging publicly available data and open source content housed in GitHub.
“Students can have a hard time with the discrete mathematics involved, so I have to connect it to real life. One of the first activities I do is have students explore Zillow.”
Since the foundation of machine learning/A.I. is probability, she has students replicate that process with way less data, a single website, and simpler math.
“In this activity, students choose a certain number of houses, and we build toward a Zestimate, or an estimate of what it is worth. Zillow gets their Zestimate through a set of calculations involving probability and error margins. Each time a house sells, Zillow must adjust their algorithm to account for the new data. This is the same thing that A.I. does – just faster and with much larger sets of information.”
Dr. Chung attributes her ability to plan such lessons to the giving nature of the computer science community. A lot of the open source content we see today is the result of computer science leaders like Coursera founder Andrew Ng, or colleges like Harvard.
Her content isn’t limited to higher education – she’s written a children’s book (AI: The Magic Box) with a few colleagues to help K-12 teachers introduce students to the foundational concepts of A.I. that she previews in her TedTalk.
There are many misconceptions around the field, but the biggest one to overcome is getting everyone to speak the same language.
“Statistics and computer science content have so much overlap, but they have different vocabularies. For example, most people think coding means python, but using sigma as sum in excel IS coding.”
For those who are A.I.-curious, she recommends starting with a few of her favorite resources:
- Dr. Rachel Chung’s Github contains excel sheets that mimic A.I.
- Learn Your Way with Google uses A.I. to customize the student’s experience with their text
- AI By Hand by Professor Tom Yeh shows the math behind artificial intelligence
- Harvard’s CS50 is one of the largest and oldest open source introductory computer science courses
- “Build with Andrew” features Coursera founder Andrew Ng and breaks down writing code in the context of A.I.
- “LLM From Scratch” by Sebastian Rascha argues anyone can create an LLM
Dr. Chung is presenting with colleagues from George Mason University in March on how to use A.I. to streamline writing drafts. You can learn more on her LinkedIn.
AUTHOR: Dr. Holly Whistler coordinates distance education for the Virginia Community College System. She is also a member of CCCOER’s Research and Impact Committee. Her decades of education experience include teaching future teachers and writers, supporting students with disabilities, and designing courses and support services for distance learners. She is a reviewer for the Online Learning Journal and AERA, with a focus on adult learners, faculty experiences, digital literacy, and online programs. She serves on her local library board and continues to advocate for the freedom to read and learn. You can contact her at hwhistler@vccs.edu..
Featured image: Artificial Neural Network with Chip, by Mike MacKenzie via Wikimedia Commons, licensed under the Creative Commons Attribution 2.0 Generic license