Stanford AI-Related Classes Have Doubled in the Past Five Years

• Stanford students say interest in generative AI has already surpassed the crypto hype.
• Enrollment in existing classes on artificial intelligence has skyrocketed, according to faculty.
• A look at the course catalog shows AI-related classes appear to have doubled in the past five years.

It’s hard to separate Stanford University from the orbit of the tech industry. And as a generative AI craze sweeps through Silicon Valley, it’s set big changes in motion at Stanford, too

Just look at Stanford’s course listings. In just five years, the number of classes related to artificial intelligence appears to have doubled. 

Every year, Stanford graduates funnel into coveted gigs at Big Tech companies, venture capital firms — or they embrace the startup grind as founders. Almost 40,000 companies have been launched by alumni and faculty since the 1930s, including giants of the tech economy, like Google, LinkedIn, and Yahoo, the university brags.

Now, since OpenAI released ChatGPT in November, interest in AI on campus has surged, more than a dozen Stanford students and faculty told Insider.

“The speed the industry is moving is very surprising to me — as a person who is very aware of the industry,” said Rishi Bommasani, a computer science doctoral student at the Stanford Center for Research on Foundation Models. And Stanford’s rush into AI? “The speed of Stanford is very surprising to me — as a person fairly familiar with Stanford.”

AI is now brunch conversation

Until recently, conversations around AI have largely been relegated to those with a technical fluency in machine learning, programming, or advanced mathematics, students told Insider. 

The current curiosity, though, has been fueled by the accessibility of generative AI. Anyone can experiment with ChatGPT, for example, and use their experience as a starting point to consider broader questions around artificial intelligence.

Now, students majoring in everything from English to economics to symbolic systems (also known as the “VC major”) told Insider that they’re grappling with the political, social, and even philosophical implications of the AI revolution.

Take, for instance, a recent conversation among a group of Stanford students: Senior Isabelle Levent and her boyfriend had a few classmates over to their dorm for brunch in March. Guests initially circled through the usual topics — the a sinkhole in the parking lot, a venture capital event someone attended, predictions on when a lake nearby might run dry. 

That’s until someone brought up a research paper that had just been published by OpenAI, the startup behind ChatGPT and its newest language model, GPT-4.

Suddenly, everyone’s faces lit up, Levent said. The group began discussing the paper, GPT-4’s capabilities, and their concerns about the safety of AI. “This isn’t typical brunch talk for all students on campus,” Levent later told Insider by email, “but AI is on a lot of people’s minds.”

‘Fast’ even for Stanford

The university has made swift moves to accommodate the burst in student interest.

Stanford is a decentralized campus where course programming is done at the departmental level, a representative for the university told Insider by email. That means it’s difficult to tell exactly how many AI-related courses were introduced over the past few quarters.

But Insider went through course catalogs from the past five years to get at least a bird’s-eye view of the increase: For the 2018-2019 school year, there were 72 courses that touched on artificial intelligence in some capacity — and just five that specifically mentioned “generative AI.”

By the 2022-2023 school year, that number jumped to 140, with 14 courses that specifically touched on “generative AI.” That means courses related to AI — at least by Insider’s look — appear to have basically doubled in the past five years.

And enrollment in Stanford’s artificial intelligence-focused computer science classes also has skyrocketed in recent months.

Over the past five years, CS 224N, a popular computer science course on natural language processing, has drawn about 400 to 500 students in the winter quarter. This year, though, more than 650 students enrolled in the class, making it the most popular computer science class, and second-most enrolled course in the quarter, according to The Stanford Daily. 

Over the past few months, the university has also added new classes on generative AI in medicine, economics, and its master’s program in liberal arts — all in departments that don’t regularly touch on the topic.

And professors are putting these courses together with record speed, too. 

Bryant Lin, a clinical professor of medicine, is teaching a class this quarter called “Generative AI and Medicine” that he said he organized in a matter of months. 

Given the size of Stanford’s course catalog, Lin said that new classes don’t land on students’ radars as quickly as existing ones, but his class filled up in less than a day with a 36-person waitlist. 

For professors who have dedicated years to studying artificial intelligence, the uptick in interest has raised a bigger question: How should instruction be expanded beyond the student pursuing a Ph.D. in computer science?

AI expert Peter Norvig began considering this question in the early 2000s when he was directing the core search algorithms group at Google and put together a course for software engineers who had little experience with machine learning. Now, as a fellow at Stanford’s Institute for Human-Centered Artificial Intelligence, Norvig is thinking about how to broaden the scope of education to the non-technical student. 

“I think, just for, you know, everyone — as a consumer and citizen — these issues are touching our society,” Norvig said. “They should know enough to understand what’s going on, to use the tools that are available to them, to take advantage of the benefits, to watch out for the potential harms, and to understand how it’s affecting society.”

Of course, the acceleration of generative AI tools like ChatGPT has also raised the question of whether coding, or essential academic skills, will become obsolete.

Siddharth Sharma, a sophomore majoring in computer science, said those doubts haven’t yet swept through campus.

“The whole point of ChatGPT is that it makes it easier to write segments of code that already exist out there in the wild,” he said. “It just means that more critical thinking is going to be necessary, you know, in the coming years for students in general.”

AI isn’t the next crypto

The frenzy is not only academic. 

Stanford’s deep ties to the venture capital world — where the generative AI boom has been a rare bright spot amid a gloomy funding landscape — have pointed students toward the tech industry’s next gold rush. 

In conversations with Insider, several students drew comparisons between the AI hype cycle and the crypto craze of years past. Their general consensus, though, was that while interest in crypto was purely monetary, the fascination with generative AI is rooted in its potential to upend and reshape society.  

As symbolic systems major Sophie Fujiwara put it via Twitter direct message: The interest in AI is greater than any other trend.

Every “entrepreneurially-minded” undergraduate is experimenting with generative AI, developing notes apps, health apps, and creating content, Levent, who hosted the brunch in March, told Insider. Graduate students studying artificial intelligence are also debating whether to leave to noble world of academic for lucrative gigs in Silicon Valley’s new promised land, Insider previously reported.  

For the undergraduates, at least, the ideas are still nascent. They aren’t yet dropping out in droves to launch AI startups.

Recent graduate, Bryan Chiang, who recently built RizzGPT, an AI-powered monocle that displays words to help users sound more charismatic everyday conversations, said that most students are just “tinkering” right now. 

But that doesn’t mean the promise of a larger disruption is on the horizon.

“You still have your Big Tech companies lying around, but they’re much more vulnerable, because the bleeding edge of AI has basically been democratized,” Chiang said, who spent time conducting research on AI models at Google and Microsoft.

He summed up the promise of AI succinctly: “You don’t have to be an academic or a professor, or someone with 10 years of expertise, to do cool stuff.”