Yann LeCun, chief AI scientist at Meta and a pioneer in the field of artificial intelligence, has shared his advice for students considering a career in computer science and AI. In an email to Business Insider, LeCun warned that simply following the minimum math requirements in a typical computer science curriculum could leave students unprepared for the rapid evolution of technology. "if you are a CS major and take the minimum required math courses for a typical CS curriculum, you might find yourself unable to adapt to major technological shifts," he said. LeCun, who teaches computer science at New York University, joked that he often comes across as a "computer science professor arguing against studying computer science" because of his strong emphasis on foundational knowledge. His core recommendation? Focus on building a deep understanding of fundamental disciplines like mathematics, physics, and electrical engineering—subjects that offer long-term value. "My advice is not to avoid CS as a major, but to take the maximum number of courses in foundational areas rather than just chasing trendy tech topics," he told Business Insider. LeCun stressed the importance of learning skills with enduring relevance. "Learn things with a long shelf life," he said during a recent appearance on the podcast "The Information Bottleneck." He highlighted the value of understanding mathematical modeling and how it connects to real-world problems—something he noted is often better taught in engineering programs. In many U.S. engineering schools, students are required to take multiple levels of calculus, as well as courses in control theory, signal processing, and other areas that are directly applicable to AI. "In computer science, you can get by with just Calculus 1. That’s not enough," LeCun pointed out. He believes these engineering-based subjects provide a stronger foundation for innovation in AI. LeCun himself did not start in computer science. He earned a degree in electrical engineering from ESIEE in Paris before completing a Ph.D. in computer science from Sorbonne Université in 1987. His background, he said, gave him a critical edge in understanding the physical and mathematical underpinnings of AI. While he acknowledges the need to learn programming, he cautions against overemphasizing "vibe coding" or relying too heavily on AI tools to write code. "Obviously, you need to learn enough computer science to program and use computers," he said. "Even as AI makes coding more efficient, you still need to understand how it works." Other industry leaders echo this sentiment. OpenAI’s Bret Taylor has emphasized that computer science is about more than just coding. Nobel laureate Geoffrey Hinton has also highlighted the lasting value of core knowledge—especially in math, statistics, probability, and linear algebra—stating that these are the skills that won’t become obsolete. As universities and students grapple with how to adapt to the rise of generative and agentic AI, LeCun’s message is clear: build a strong foundation. The most valuable skills are not the latest tools or frameworks, but the deep, enduring knowledge that allows you to understand, innovate, and lead in a changing world.