English major

An informal but increasingly accurate piece of vocabulary describing one of the more surprising labor-market shifts of the 2024-2026 agentic-AI transition.

In one sentence

An “English major” — used in the agentic-AI context — is the kind of person who turns out to be surprisingly good at directing AI coding agents because the work has quietly become a writing-and-judgement job rather than a syntax-memorization one, and who is now interviewing for software roles at companies that, three years ago, would have screened them out at the resume stage.

Why this term exists

For roughly forty years, the cultural assumption in the technology industry was that coding and writing were two different skills exercised by two different kinds of people. Software engineering was held to be a STEM discipline; clear writing was a separate, softer competency, often dismissed as nice-to-have.

Then agentic coding tools arrived. The bottleneck shifted. The question stopped being can you remember the syntax for a hash map in this language? and became can you specify what you want clearly enough that the model gets it right? The first question is a memorization problem. The second is a writing problem. The second turns out to be much harder, much rarer, and much more valuable.

The English-major phenomenon is the labor-market reflection of that shift. Companies that started their agentic-AI rollouts in 2024 quickly noticed that the people getting the best results from the new tools were not always the same people who had been getting the best results from the old tools. A non-trivial number of those over-performers had liberal-arts degrees, journalism backgrounds, or law-school training — disciplines that train clear specification under ambiguity as the central skill.

What it actually looks like

The “English major” pattern in practice:

Working example

In 2025-2026, several large engineering organizations reported that their highest-leverage agentic-coding power users were often people who had originally been hired for technical-writing, product-management, or developer-relations roles. Not exclusively — strong coders who could also write remained extraordinary. But the floor of the new top performer pool included people the prior screening process would have rejected. By 2026, several public job listings explicitly stated “Liberal-arts background welcome; ability to specify clearly is the core skill.” This was not a charitable inclusion. It was an observation about who was getting the work done.

Why this matters in a teaching context

For a BBA or MBA classroom, the English-major phenomenon is one of the cleanest cases of a value-chain disruption inverting the prior skills hierarchy that students are likely to encounter in their careers. Yesterday’s bottleneck (syntax) is automated. Yesterday’s nice-to-have (clear writing) is the new bottleneck. The labor market is repricing accordingly.

For the Isenberg context specifically: the management student who took the writing-intensive electives, who learned to write tight memos, and who can specify ambiguous business problems unambiguously is — in 2026 — better positioned for many AI-augmented roles than they would have realized two years ago. Worth saying out loud in advising conversations.

A second framing: the English-major shift is what zhengming (the rectification of names) looks like when an industry has been mis-naming its bottleneck for forty years. The bottleneck was never syntax. It was always specification. The hiring practices that mistook one for the other are now correcting themselves.

The prophetic case: William Gibson

The English-major phenomenon has a forty-year precedent that nobody in the technology industry adequately absorbed the first time.

William Gibson studied English at the University of British Columbia. He was not a programmer. He had no technical background in computing. In 1984, he published Neuromancer — the novel that coined the word cyberspace and described, with unnerving precision, a networked information world that did not yet exist.

His explanation for how he got there is worth reading slowly. He was not modeling technical systems. He was watching children at video arcades:

“I could see in the physical intensity of their postures how rapt the kids were… It was like a feedback loop, with photons coming off the screens into the kids’ eyes, neurons moving through their bodies, and electrons moving through the video game. These kids obviously believed in the space these games projected.”

That observation — extracted from a sidewalk, by a novelist paying attention to the wrong people in the right room — produced the conceptual vocabulary that the internet industry spent the next decade building toward.

But Gibson has said something more precise about the act of coining the word itself. Cyberspace, he has explained, was invented as an empty vessel — a word that sounded like the thing he was trying to name, before he knew what the thing was. He needed a container. He built one out of sound and feeling. And then reality, over the following decade, poured itself in.

In his short prose piece Academy Leader (1991), Gibson described the act in his own voice:

“Assembled word cyberspace from small and readily available components of language. Neologic spasm: the primal act of pop poetics. Preceded any concept whatever. Slick and hollow — awaiting received meaning.”

“All I did: folded words as taught. Now other words accrete in the interstices.”

This is naming as a generative act, not a descriptive one. The word did not describe an existing thing — it preceded any concept whatever. It created the attractor around which an existing thing could organize, and meaning accreted in the interstices over the following decade. The internet did not make “cyberspace” accurate; “cyberspace” helped make the internet legible. An English major understood this before the engineers did, because the engineers were building the vessel and Gibson was already thinking about what it would hold.

The underlying capacity is imagination — not in the casual sense of creativity or whimsy, but in the precise sense: the ability to hold something that does not yet exist with enough clarity to name it. STEM disciplines, at their best, train rigorous analysis of what is. Imagination works on what isn’t yet. Gibson’s method — observing the human signal in the technological noise and naming what he saw with sufficient precision — is imagination applied as a technical instrument. The name creates gravity. Reality organizes around it.

This is why the English-major phenomenon is not simply a labor-market curiosity. It points at something deeper: the skills that the technology industry systematically undervalued for forty years are exactly the skills that produce the conceptual vocabulary a field needs to understand itself. Gibson gave the internet its name before the internet existed. That is not a coincidence of talent. It is what trained imagination, applied seriously, does.

The same method runs through his essay “Doing Television” (published in Tesseracts 3, the Canadian science fiction anthology, 1990; a slightly longer version titled “Darwin” ran in The Face and Spin the same year). In it, a girl named Kelsey, in a room in Darwin, Australia, buys a cassette tape from a man on the street — and Gibson, watching, begins to extract what the transaction means about technology, value, and culture at the edge of the global system. The scene has nothing to do with computers. It has everything to do with how technological objects move through human worlds in ways their designers did not intend.

“The street finds its own uses for things.” That line, from the same period, became one of the defining observations of the digital age. An English major wrote it.

Forty years later, in Agency (2020), Gibson described the architecture of agentic AI — agents with scoped powers, oversight structures, the authority to terminate, the Lowbeer Question — with enough precision that practitioners building real systems in 2026 have found it useful as a design document. Not because he predicted the technology. Because he specified the human relationships around it precisely enough that the map preceded the territory.

This is the English-major method at its fullest extension: find the human signal in the technological noise, name it with enough precision that the name becomes load-bearing, and trust that the specification will outlast the specific tools it was written about.

Gibson is the existence proof. The phenomenon that is now repricing liberal-arts graduates in engineering organizations is the same phenomenon that produced Neuromancer in 1984 and Agency in 2020. He was just doing it forty years early, without the agentic tools to assist him.

Trade-offs

Related entries: Naming, Tool, Agent, The Lowbeer Question, forthcoming Bathrobe coding.

Return to Dictionary All Entries (A–Z)