Whither Computing?

The emergence of GenAI-based coding agents is changing software development as we know it.

A new team will assume the leadership of ACM on July 1, 2026, following the current general election and search for a new chief executive officer. I believe this team will have to grapple with existential questions about the future of computing as a science and a profession.

In a meeting I recently had with 20 Ph.D. students in a department I visited, it became clear they were asking themselves if they had made a good decision to pursue a doctorate in computer science (CS). They were closely watching the research-funding crisis.^a They know that research grants fund their graduate stipends, so their worry is quite personal. They have also read the August 2025 article in The New York Times with the sensational title “Goodbye, $165,000 Tech Jobs. Student Coders Seek Work at Chipotle.”^b Software-development job postings on Indeed.com in the U.S. peaked^c around mid-2022 (the COVID peak) and have since flattened. So, the students expect CS undergraduate enrollment to drop, and they know that shrinking undergraduate enrollments means fewer teaching assistantships and fewer academic job openings.

One may be tempted to believe we are just facing a cyclical decline. We can recall the “Image Crisis” of the early 2000s. The computing field went through a perfect storm in those years: the dot-com and telecom crashes, the offshoring scare, and a research-funding crisis. After its glamour phase in the late 1990s, the field seems to have lost its luster. It resulted in a precipitous drop in North American enrollments in undergraduate CS programs. The enrollment plunge led to the establishment of the ACM Job Migration Task Force. The 2006 report^d issued by the Task Force was optimistic on the future of computing as a profession, and the trajectory of computing over the past 20 years affirmed that optimism. Computing today seems to be booming, with the market capitalization of Big Tech approaching $20T.

But the decline in the demand for entry-level software developers is not just a normal post-COVID cyclical decline. The emergence of GenAI-based coding agents is changing software development as we know it. Senior developers tell me gushingly how the vibe-coding tools increase their productivity significantly. But how does one become a senior software developer if positions for junior developers are disappearing? In fact, tech companies are laying off^e developers by the tens of thousands.

We are accustomed to computing disrupting other fields; think, for example, of travel agents around the year 2000. But now we have disrupted ourselves, as no one knows the future of the software-development labor force. In fact, CS is the most disrupted discipline on campus these days. All disciplines face the realization that one of the most basic educational tools, homework, is now obsolete, as we must accept that students will use GenAI tools to handle homework. But CS departments must also grapple with the fact that the most fundamental skill we used to teach, which is coding (remember “Coding for All”?), may not be a viable skill at all.

Beyond the prosaic concerns about funding, jobs, and the like, the students expressed a deeper angst about the discipline they have chosen to specialize in. They are not out of touch with society at large, and they see how computing has been subsumed by AI (no one says “artificial intelligence”), and AI means LLMs. Half of U.S. adults say^f the increased use of AI in daily life makes them feel more concerned than excited. If the image crisis of the early 2000s was about a profession with diminishing prospects, the image crisis of 2026 is about an industry that is viewed, counterphrasing an old Google dictum, as one that has become “evil.” Another, more recent article^g in The New York Times asked, “Can an A.I. Company Ever Be Good?” Cory Doctorow has coined the phrase “enshittification” to describe the declining quality of tech products. But we now seem to have moved from enshittification to “envillainization.” Deplorably, but not shockingly, Sam Altman, who embodies AI to many people, was targeted in two violent attacks last April.

ACM’s next president and CEO will not have a magic wand to bring back the good old days. But if ACM is about “advancing computing as a science and profession,” then we need to engage in a deep conversation (modeled after the 2005–2006 Task Force) about what this phrase means today for education, research, the profession, and ACM, and about how to truly advance computing as a science and profession.