Yan LeCun said that if computer science majors don't use their time wisely, they may find their degree is meaningless.
“If you are a CS major and have taken the bare minimum math courses required in a typical CS curriculum, you may not be able to adapt to major technological changes,” LeCun said in an email to Business Insider.
LeCun, who teaches computer science at New York University, said in a recent podcast appearance that he joked that he was “a computer science professor who is against studying computer science,” based on his arguments about where students should focus their time.
“My recommendation is not to avoid CS as a major, but to maximize the fundamentals (math, physics, EE courses, etc.) instead of taking courses on trendy technology every day,” he told Business Insider.
The former chief AI scientist at Meta said he advises students to “learn things that have a long shelf life.” Depending on your computer science program, not all of these skills may be incorporated into your degree.
“What we need to do is learn the basics of mathematics, modeling, and mathematics that can be connected to reality,” LeCun said on the podcast The Information Bottleneck. “In some schools, you tend to study this kind of thing in engineering fields related to computer science, but also things like electrical engineering and mechanical engineering.”
Universities and computer science students continue to grapple with how to adapt their programs to the age of increasingly generative and agentic AI. Earlier this year, Hany Farid, a professor at the University of California, Berkeley, described the struggles students face in finding jobs, comparing them to the “field operations” of former graduates.
Leaders in the field, including OpenAI's Brett Taylor, emphasize that computer science is more than just learning to code. Others, including Nobel laureate Jeffrey Hinton, emphasize that learning critical thinking is key to staying ahead of advances in AI.
“Skills that are always valuable, like knowledge of mathematics, statistics, probability theory, knowledge of linear algebra, etc., are always valuable,” Hinton recently told Business Insider. “It's not knowledge that disappears.”
LeCun jokingly pointed out that he didn't study CS at first. He studied electrical engineering at ESIEE in Paris and later earned his Ph.D. LeCun said some CS schools align with engineering programs, which tend to require more advanced math.
“In engineering, in America you learn basic calculus 1, 2, and 3, right?” he said. Calculus 1 is enough for computer science. That’s not enough, right?”
Engineering also exposes students to concepts such as control theory and signal processing, which LeCun said are “very useful for things like AI.”
All of this doesn't mean you should abandon basic programming, LeCun says. Vibe coding is great, but it's no substitute for basic knowledge.
“It's clear that you need to learn a lot about computer science to program and use computers,” he says. “AI can help make programming more efficient, but you still need to know how to do it.”
