Even as AI progresses, coders aren’t doomed.
It’s a weird time to be studying computer science. Recent grads have a higher unemployment rate than those in just about every other major—yes, even philosophy. The internet is littered with rants from newly minted programmers who can’t find work. On one such YouTube video, the top comment reads: “Your first mistake is not being born earlier.” Students, meanwhile, are fleeing the field. Undergraduate enrollment in computer science dipped by more than 8 percent last year, representing the largest absolute decline across any major in several years. The falloff at the graduate level—14 percent—was even more severe.
Learning to code was supposed to be a ticket to a good tech job. It wasn’t just Silicon Valley that spread the gospel of computer science: “Support tha american dream n make coding available to EVERYONE!!” Snoop Dogg once tweeted. Now the decision to major in CS is more complicated. Nowhere has AI refashioned work as dramatically as it has for programmers. Coding bots have become much more powerful over the past few years, and they excel at precisely the kind of programming that might previously have been delegated to entry-level workers. An Anthropic co-founder, Jack Clark, recently warned that “the value of more junior people is a bit more dubious,” as some 90 percent of the company’s new code is apparently now AI-generated.
The popular narrative around CS has flipped to such a degree that some Silicon Valley insiders are now actively discouraging people against the major. John Coogan, a co-host of TBPN, a popular tech-news podcast, recently asked if it would be a “contrarian move” to study computer science “at a time when coding jobs are going away.” But studying computer science is not contrarian, and the major’s waning relevance has been overstated.
It’s true that the work situation is more dicey than it once was. “Forget Python, study Plato,” The Economist advised students last week. But although the unemployment rate for new CS grads is spiking, they have a relatively low rate of underemployment—that is, comparatively few are working in jobs that don’t usually require a college degree. (Consider that nearly half of philosophy majors are underemployed.) When it comes to wages, new computer-science grads are also still significantly outearning their peers. One explanation for why CS majors have such high unemployment rates is that they may be less likely to settle for lower-paid roles. If you’re optimizing for earnings, trading software for Socrates might not make so much sense after all.
None of this is to dismiss the AI threat to software jobs. The aforementioned employment data tracks students who graduated in 2024. AI has improved significantly since then, and the capabilities are likely to continue to increase, allowing bots to take on more sophisticated work. But the decline of manual programming—that is, writing code by hand—doesn’t obviate the need for computer scientists. Even as AI tools become more powerful, leveraging bots to build reliable and secure software still takes training and expertise. With the AI revolution in full swing, we are hurtling toward a future in which even more of the global economy is mixed up with the software industry. If anything, the AI-ification of work seems likely to require more people who understand computer systems at a deep level. Across the tech industry, demand for mid- and senior-career engineers is rising. The trouble, then, is how to adjust today’s computer-science programs to equip students for work when the field is changing so fast—especially when entry-level coding jobs that once were guaranteed are now far less certain.
“I don’t know where the world is going,” Michael Hilton, a computer scientist at Carnegie Mellon University, told me, “but I know the things I taught three years ago are not the right things to teach today.” As bots have become more capable, Hilton keeps updating his curriculum—he encourages students to use AI for coding. Other professors are moving in the opposite direction. Valerie Barr, a computer scientist at Bard College, told me that in her introductory class, coursework is now mostly done on paper. “I’m back to how I taught in the 1980s, when we didn’t have laptops and there was one computer lab for the whole campus,” she said. Barr believes that students who learn coding fundamentals the old-fashioned way will be the ones to come out ahead. “You cannot make effective use of AI tools if you don’t know something about what you’re asking the tools to do,” she said. In much the same way, grade schoolers learn how to do basic algebra by hand before they are allowed to use calculators.
The split over whether to embrace coding tools points to a larger divide in the discipline: Is studying computer science about training students to be good software developers, or teaching them the computational theory that underpins the field? As coding becomes automated, we might see a further fracturing between the two domains. On the theory side, the AI boom has put a premium on highly skilled researchers with a deep understanding of machine learning. Future students may enroll in new AI-related majors that take the conventional CS major and then layer in more specialized AI training. Such programs already exist at several colleges: MIT introduced an AI major in 2022, and it’s already become the second-most-popular major on campus—behind computer science. And some students who are interested in CS for its own sake will still go deep in other non-AI subfields, such as cryptography. Today’s AI boom is possible only because people pursued neural networks when they were uncool.
At the same time, new courses could offer students an introduction to software development without the theoretical baggage and proof-writing they might have otherwise had to wade through. Geoffrey Challen, a computer scientist at the University of Illinois at Urbana-Champaign, plans to offer a new course this fall in which he will teach students to develop software “without writing, reading, debugging, or viewing a single line of code,” he told me. Northwestern is also slated to offer an “entry-level creative coding” class for students without technical backgrounds. For all the talk of AI-literacy programs that teach students how to use chatbots, the real innovation might be in developing courses that train students in basic software-development skills. Most colleges require introductory writing courses because it’s understood that clear written communication is an important cross-disciplinary skill—even for students who plan to study physics or math. Classes that teach students how to use AI coding tools could become commonplace, providing students of all backgrounds with a baseline software-engineering skill set.
The days of computer-science grads being all but guaranteed cushy tech jobs may be coming to an end, and the next few years will almost certainly be tumultuous as the job market continues to adjust. But we’re on the precipice of a new era when learning to develop software will be easier than ever, opening the door to students who might not otherwise have chosen to study computing. Perhaps a new golden age of CS education has only just begun.