Modern CS: The Age of Minors and Certificates

It seems that every other week or so, somebody from one of the frontier AI labs would drop an apocalyptic prediction on social media and interviews regarding how AI is (quickly becoming) capable of replacing human in different types of jobs. While these kind of statements should be taken with a large spoon of salt, a study from March 2026 from Anthropic found that while there are jobs that are significantly more exposed to AI integration, there is no impact on unemployment rates for workers in the most exposed occupations, although “there’s tentative evidence that hiring into those professions has slowed slightly for workers aged 22-25”.

Rather than focusing exclusively on whether jobs are disappearing, in this essay, I want to echo the sentiment of a recently blog entry by Andrew Ng that it may be more productive to examine how the structure of computing work itself is changing. The article by Davenport and Paredes calls for a more analytical view into the matter where we focus not on the professions themselves but the individual tasks that make up these professions. How will this view impact the administrative structure of traditional CS education perspectives regarding alternative pathways such as minors and certificates?

From Jobs to Tasks

From my own perspective, the AI-impacts-on-jobs discussion is a very interesting one. As an educator in the CS field, I definitely want to say that no, AI cannot replace computer scientists. At the same time, as I take my educator hat off and put my engineering hat on, I am elated by how much AI has enabled me to accomplish. As always, the right answer is somewhere in the middle.

This creates a very different computing reality from the one often presented online.

For these organizations, the goal is not to eliminate software developers entirely. The goal is to make existing workers more productive. A small company may still employ developers, but now those developers are expected to work across a broader operational surface area. They may maintain databases, automate workflows, integrate AI services, clean internal data pipelines, manage cloud infrastructure, and communicate directly with nontechnical departments. In many cases, the most valuable employee will not be the person who knows the most obscure algorithmic trick, but the person who understands both the operational domain and enough computing to improve it.

This is also why I increasingly find the term “AI engineer” somewhat meaningless in isolation. What exactly is being engineered? AI infrastructure for a bank? A manufacturing workflow? A medical documentation system? A logistics platform? A fraud detection pipeline? A local retail business trying to automate inventory management? The phrase only becomes meaningful once it is attached to a real domain problem.

I suspect this is where computing education is heading over the next decade. Computing itself is becoming less isolated as a profession and more embedded into every other profession. In some sense, computer science may slowly become more like mathematics. Not because advanced computer science disappears, but because baseline computational literacy becomes necessary for almost everyone.

There was a time when driving was considered a specialized skill. Only a relatively small number of people knew how to operate vehicles safely and effectively. Today, driving is largely treated as a basic societal competency. Yet professional driving still exists. There are race car drivers, truck drivers, heavy equipment operators, and logistics specialists whose expertise goes far beyond ordinary commuting.

I increasingly wonder if computing is moving toward a similar structure.

Perhaps we should stop imagining that every computing student is training to become an elite software engineer at a major technology company. Some students certainly will. Those are the equivalent of professional racers or specialized operators. But many students may instead need enough computing knowledge to function effectively inside biology labs, hospitals, manufacturing companies, banks, government agencies, media organizations, or small businesses. They are not necessarily building operating systems or distributed AI platforms. They are using computational thinking to solve practical domain problems.

From a computer science educator perspective, this raises uncomfortable but important questions. What exactly are we training students to become? Are we training specialists? Generalists? Infrastructure builders? Domain translators? Are we preparing students only for large software companies, or are we preparing them for the broader economy that actually exists around us?

At regional public universities, these questions matter enormously. Many of our students will not work at frontier AI labs. Many will remain in regional industries. Some will become hybrid professionals who combine computing with psychology, healthcare, manufacturing, finance, biology, communication, or education. If this is the direction the economy is moving, then our curricular structures may need to change alongside it.

I suspect we will continue seeing an expansion of computing minors, certificates, and interdisciplinary programs. Not because depth no longer matters, but because breadth is becoming economically valuable in ways that universities have historically underestimated. A biology student with practical AI and data skills may become extraordinarily valuable in a biomedical environment. A psychology student who understands LLM workflows and data infrastructure may contribute meaningfully to clinical training systems. A manufacturing student with automation and cloud knowledge may transform operational processes without ever identifying primarily as a software engineer.

Ironically, this may also reduce some of the prestige mythology surrounding computing itself. For years, computer science has often been culturally framed as a rarefied profession reserved for highly technical specialists. But as computing tools become more accessible, the discipline may begin to resemble other infrastructural literacies that quietly underpin modern society. This does not make the field less important. If anything, it makes it more important. Infrastructure becomes invisible precisely because society depends on it everywhere.

At the same time, I do not believe deep expertise disappears. Advanced systems engineering, distributed computing, cybersecurity, compiler design, embedded systems, and machine learning research will likely become even more specialized. We will still need highly trained experts building the computational foundations that everyone else relies on. But the surrounding ecosystem may expand dramatically outward.

Perhaps the real AI transformation is not that software engineers disappear. Perhaps it is that computing itself stops belonging exclusively to software engineers.