Yann LeCun's advice for young students wanting to go into AI

Yann LeCun, a prominent figure in artificial intelligence and a computer science professor at NYU, has shared critical advice for students aspiring to enter the field of AI. In his insights, he warns that merely completing the minimum math requirements in a computer science curriculum may hinder their adaptability to rapid technological advancements. In a recent podcast discussion, LeCun humorously remarked on his advocacy for a robust academic foundation, saying he feels like a computer science professor who is cautioning against a narrow focus on computer science itself. He encourages students to prioritize foundational courses in mathematics, physics, and electrical engineering over more transient tech trends. "Learn skills with lasting relevance," LeCun emphasized. He pointed out that many computer science programs may not adequately cover these essential skills. During his conversation on "The Information Bottleneck" podcast, he highlighted the importance of understanding mathematical concepts that are applicable to real-world scenarios, often found in engineering disciplines. As universities adapt to the rise of generative AI and its capabilities, challenges persist for students in securing employment. Hany Farid, a professor at UC Berkeley, noted that job opportunities are not as abundant as they once were, contrasting the experiences of current graduates with those of the past. LeCun, along with other industry leaders like OpenAI's Bret Taylor, stresses that a successful computer scientist must possess more than just coding skills. Nobel Laureate Geoffrey Hinton echoes this sentiment, advocating for the necessity of critical thinking in navigating the evolving landscape of AI. Hinton stated, "Skills such as mathematics, statistics, and probability theory will always hold value. Knowledge in linear algebra is also timeless." LeCun shared a personal anecdote, revealing that his own academic journey began in electrical engineering at ESIEE in Paris, culminating in a Ph.D. from Sorbonne Université in 1987. He noted that engineering programs in the U.S. typically require a more comprehensive mathematical education, including multiple calculus courses, which provide a solid foundation for students. He underscored that while programming basics are important, they should not overshadow the necessity of fundamental knowledge. "You need to learn enough computer science to program and utilize computers effectively," he stated, acknowledging that while AI tools can enhance programming efficiency, foundational skills remain essential.