Artificial intelligence degrees are no longer just for the universities that teach tech geniuses.
Only five schools offered A.I. majors in 2021. Now, universities are setting up programs so quickly that researchers are struggling to track them. At least 74 A.I. majors and 89 minors are available on American campuses, according to Northeastern University’s Center for Inclusive Computing.
At least another dozen schools, many of them far from Silicon Valley, are poised to debut majors this year, reflecting the frenzy around the technology and academia’s urgent ambition to be seen as essential in the A.I. age.
The idea is to keep schools and students alike competitive as A.I. reshapes the global economy. The new programs, though, vary widely in their details, with some emphasizing the inner workings of A.I. and many others more focused on how to use it. And it is unclear how students who earn the degrees will fare as companies recalibrate.
“Some call it a bubble. Maybe it is,” said Uzezi Olorunmola, who is seeking his doctorate in A.I. at the University of North Dakota, the flagship school in a sparsely populated state that will soon have two universities offering degrees in A.I.
“But I think it’s pretty much here to stay, and the earlier you not only get with the program but also know how to use A.I. or use A.I. applications, I think it’s better.”
Sitting at a conference table with other students in Grand Forks, he acknowledged with a grin, “We’re basically the test subjects.”
Glitz or Substance?
Academia’s stampede toward A.I. programs is challenging the perception that higher education is plodding. It is also inviting questions about whether colleges are sacrificing quality in a rush for relevance.
“We have to be careful: Is it glitz or is it substance?” said Andrew Armacost, the president at North Dakota, which had eight students enrolled this past academic year in its new A.I. doctoral program.
The answer varies from school to school and is often discernible by comparing degree requirements. Many colleges’ A.I. program curriculums overlap substantially with those for computer science. Sometimes, only a handful of new courses distinguish the degrees.
And A.I. degrees can vary sharply in their goals.
Several universities are steeping their programs in theory, intending to turn out graduates who can do the under-the-hood work that powers A.I. These include places like Carnegie Mellon in Pittsburgh, which in 2018 became the first American university to offer an A.I. degree. Thirty-three courses are now approved for the major.
Reid Simmons, a Carnegie Mellon computer science professor who directs the A.I. major there, suggested tech companies looking for the next generation of A.I. experts would focus on elite institutions with track records, just as an elite law firm would hire lawyers.
“The people who are looking for lawyers know the difference between a Harvard law degree and some other law degree,” he said.
Some schools regard the new programs as boons for students who otherwise would have been computer science majors as graduates struggle to find jobs and universities report plunging or plateauing enrollment in the discipline.
At many schools, the degrees are often meant to appeal to students who want more training in A.I., whether or not it is central to their discipline.
They are not carbon copies of computer science programs.
At North Dakota, for example, A.I. doctoral students complete fewer dissertation research credit hours than their computer science counterparts. And while the A.I. curriculum requires classes in subjects such as applied machine learning, the menu of electives includes topics like data science ethics, bioinformatics and quantum and computational chemistry.
Some schools are developing new classes and hiring professors. At others, administrators have told trustees and regents that the programs will barely cost anything because they use current courses and faculty members.
Professors insist that the new options are not simply glorified lessons in chatbot prompts. At North Dakota State University, which now offers a minor in A.I. and will offer a major this year, the curriculum will include newly designed courses on topics like A.I.’s social consequences and data security.
“We don’t just teach large language models, and I don’t think we would be doing the students a service if we really limited ourselves to that,” said Anne Denton, the undergraduate program coordinator in the computer science department.
Growing Student Demand
In a submission to the State Board of Higher Education this year, North Dakota State said that demand for a full-blown A.I. degree was “strong and growing” and that there was “very limited regional supply.”
It predicted that a bachelor’s program would have 60 students within five years.
In Grand Forks, University of North Dakota leaders and professors believe that the A.I. fervor is not contained to bachelor’s degrees and that people already in the work force want — or need — to learn about the technology.
“They contact me, and they say, ‘Listen, in my job, I have to work with all of these data analytics tasks,’ and they have been looking for programs that will give them some more foundational knowledge,” said Emanuel S. Grant, the graduate program director for the School of Electrical Engineering and Computer Science.
Administrators said they had admitted another 10 students for the fall semester.
Will companies trust an A.I. degree?
Whatever the academic focus, many administrators see A.I. programs as ways to show new value as universities face accusations that they too often offer costly, depreciating degrees and are slow to pivot.
“Academia does take its time to think things through,” said Ed Seidel, president of the University of Wyoming, which offers an A.I. master’s program.
“But,” he added, “I think everyone is recognizing now that we really have got to be preparing our students for this.”
Dr. Seidel and others acknowledged that the new degrees could attract prospective students. He also had a warning: Embracing A.I., which polling shows stirs unease among Americans, may not help higher education as it tries to win back public trust.
A.I., Dr. Seidel said, “doesn’t necessarily win an argument that you’re relevant if people are scared of it.”
How employers will judge A.I. degrees remains an open question. The reputations of longstanding programs in other disciplines, academic leaders said, could shape how people respond to new ones.
“I’m not worried about people who are rebranding or adding a few courses to what are already quality C.S. degrees,” said Charles Isbell, a computer scientist and the chancellor of the University of Illinois Urbana-Champaign, which is not offering a stand-alone A.I. degree. “I would worry about people who are just trying to shift into this space.”
Administrators at schools like North Dakota, which first offered a computer science major in 1971, understand the doubts — in part because they wrestled with them themselves. But they were persuaded, they said, that an A.I. program could mint graduates better prepared for the work force.
They also suggested that waiting could have been competitively perilous.
“If we’re perceived as out of the game, it’s going to be hard to get our name in the game,” said Ryan Adams, the dean of the college that includes the A.I. program.
That mind-set worries some, who question whether the degrees will have staying power, and whether the programs will prove another example of A.I. stirring too much excitement.
Lisa Meeden, a Swarthmore College professor who has worked on A.I. for more than 30 years, wondered whether the programs are examples of “A.I. perpetuating its own hype.”
Many professors and administrators, though, are undeterred. Moh Rasouli, who directs North Dakota’s School of Electrical Engineering and Computer Science, noted that his own graduate school professors had dismissed A.I. as a fad.
Students are similarly unbothered.
“We have to keep up and we have to grow with the fast pace that this thing is going,” said Christianah Jemiyo, who wants to use her doctorate to work on health care.
As she and her counterparts approached the end of their program’s inaugural year, their professors were already thinking about whether an undergraduate offering might be next.
