The 'godfather of AI' says now is not the time to give up on your computer science degree, even if AI changes the coding industry.
“A lot of people think that a CS degree is just programming or something,” AI pioneer Jeffrey Hinton told Business Insider. “Obviously, just being a competent mid-level programmer doesn’t make your career last that long, because AI can do that.”
Hinton said he thinks the value of a CS degree goes beyond just coding, which is why he thinks “a CS degree will be valuable for quite some time.”
Hinton's views echo those of other leading figures in the AI and technology fields who argue that it is premature to declare CS degrees a victim of the AI boom, even though agentic AI is disrupting the job market for this once highly sought-after degree.
OpenAI Chairman Brett Taylor, who earned bachelor's and master's degrees in computer science from Stanford University, said earlier this year that a CS degree is “extremely valuable.”
“There's a lot more to coding than just writing code,” he said. “Computer science is a great major to learn systems thinking.”
That doesn't mean CS programs shouldn't adapt.
Sameer Samat, Google's head of Android, previously told Business Insider that CS needs to be restructured around “the science of problem-solving, in my opinion.”
And Hany Farid, a professor at the University of California, Berkeley, said the best jobs for CS graduates right now aren't among the “usual suspects in Silicon Valley.”
“I have always maintained that the most exciting applications of computer science are not at Facebook, Google, or Amazon, but at the intersection of computing and other fields such as computational drug discovery, media imaging, computational neuroscience, computational finance, the digital humanities, including art and music, computational social science, and policy,” Farid said in September.
Hinton's advice to middle and high school students
Hinton also strongly believes in the benefits that young students can gain from learning to code, as AI models become increasingly adept at vibe coding. This is another point of contention for some in the technology industry.
“Not so long ago, it was a kind of good intellectual activity and it helped you get a job. It's still a good intellectual activity,” he said.
Hinton said learning to code is similar to learning Latin as part of a broader humanities education.
“Even if you don’t end up having AI do all the coding, I think learning to code is very beneficial,” he said. “I think learning to code is a little like learning Latin. If you major in something like the humanities, you'll never speak Latin, but it's still useful to learn Latin.”
Overall, Hinton's advice for students aiming to become high-level AI researchers and engineers is to focus on developing critical thinking skills rather than specific skills that could be replaced by AI.
“Skills such as knowledge of mathematics, statistics and probability theory are always valuable, knowledge such as linear algebra is always valuable,” he said. “It's not knowledge that disappears.”
