Building The Future: Inside Yale's Unique Large Language Model AI Course

In Yale SOM’s Large Language Models: Theory and Applications course, students spend about six class sessions building up an understanding of how ChatGPT is built, then they put what they’ve learned to the test in the form of a project. Courtesy photos

Some schools are steadily and cautiously integrating AI into their curricula. Others are moving full speed ahead.

Yale School of Management is offering a unique course that teaches the ins and outs of generative AI. Large Language Models: Theory and Applications is unique in the sense that students aren’t just learning how to use this technology – they’re also building it themselves.

The course was launched this year. Poets&Quants spoke with the two professors who pioneered and co-teach it: Associate Dean Kyle Jensen, who leads all SOM entrepreneurship programs and programming for Yale, and K. Sudhir, James L. Frank ’32 professor of marketing, private enterprise and management.

“This course gives students a leg up in the workforce because they can point to an amazing AI project that they built themselves, and I really think that is unusual,” Jensen says. “When I look around at other business schools, I really don’t see a class like this. We are teaching something that would ordinarily be in a data science department – and also teaching the programming side. It’s just very special,” he says.

INSPIRATION AND TIMING

Jensen teaches 6 sessions on technology, and Sudhir teaches six on applications – then MBAs have a final project presentation. They use what they’ve learned and write software and use the models to do things like natural language driven customer interaction, customer research, automated data analysis, analysis of company financials.

“What truly delighted me was not only the initial interest among the students — the amount of points they bid on this course was way beyond anything I have seen in the past — but also the sustained hunger for knowledge that the students showed throughout the course,” Sudhir says.

As businesses increasingly incorporate AI, Jensen says there will be a huge need to fill all types of roles. He hopes Yale SOM MBAs will be equipped to fill many of them.

“Sudhir and I really felt like large language models like ChatGPT and similar types of generative AI techniques are having a massive effect on humanity, the business world, and society at large,” he says. “We thought we really need to be teaching this to our students because every company in the world is going to be thinking about what their AI strategy is, especially in regards to large language models.”

“My philosophy in this course was guided by my belief that when managers understand the ‘why’ and ‘how’ behind Large Language Models and Generative AI technologies, they can envision richer possibilities for their businesses,” Sudhir says.

LEARNING AND LIVING WITH AI

This course has many prerequisites – students must have calculus, linear algebra, and programming under their belt. Yet the course filled to a total of 40 students, its founders target for the inaugural cohort.

“We crafted assignments that help them understand and build on the technologies. The projects then allow them to combine their newly learnt knowledge with their own creativity,” says Sudhir. “Our students don’t simply learn about generative AI; they live it. With homework and projects, they can experiment, and envision novel innovations and solutions. It is a great testament to the power of hands-on learning even for MBA students who are more likely to be managing AI projects or guiding the strategic thinking around this technology.”

Students spend about six class sessions building up an understanding of how ChatGPT is built, then they put what they’ve learned to the test in the form of a project. Some are creating programs that monitor cryptocurrency, programs that are helping faculty draft patents, and customer service bots, but the project that excites Jensen the most is the group that is using models to create assistants and companions for the elderly.

“It’s not every day you see the overlap of emerging technology in conjunction with the elderly,” he says. “Students have been visiting the local retirement homes and developing products in collaboration with the residents there. This speaks really strongly to our school’s mission to create leaders for business and society – so for me, this project is the most special.”

Past this semester, Jensen and Sudhir expect to be teaching this course for years to come. Jensen says there is a lot of student demand.

“I’m overjoyed that the students are able to have this experience while they’re here – they’re learning about this world-changing technology right now as news about large language models and AI unfolds daily,” he says.