Asimov, programming and the meta ladder - Eli Bendersky's website

The Meta-Profession Problem

Isaac Asimov’s 1957 novella Profession contains one of the most prescient observations about automation ever committed to pulp. It is worth sitting with, because the story’s central conceit maps almost one-to-one onto the modern debate about AI coding assistants.

Spoiler warning: key plot details follow.

In Asimov’s far-future society, nobody learns a trade the hard way. At a certain age, every person’s brain is analyzed, and the knowledge for whichever occupation suits them best is uploaded directly via electrodes and training cassettes. The most capable practitioners are then funneled into prestigious “Class A” assignments on outer worlds. The protagonist, George Platen, dreams of receiving this education and securing a desirable posting. Instead, his assessment finds him unfit for any profession, and he is sent to a facility where he is free to spend his days however he likes—including reading books.

The twist is that George eventually discovers, on his own, that someone must design those cassettes, push technology forward, and continually revise training material to keep pace. He was not uneducable; he was selected for the meta-profession—the job of building the systems that make other jobs obsolete.

This is the essence of the meta ladder, and it has only become more relevant in the decades since. Most occupations are susceptible to automation, but the act of automating them simply creates a new occupation: the one that builds the automation. That chain continues upward, at least until the technological singularity.

From No-Code to Copilot

Programmers have heard this promise before. “No-code” tools would supposedly eliminate the need for developers; instead, demand shifted sideways and upward. The hype around AI pair programmers such as GitHub Copilot deserves the same measured reaction. These are genuinely useful tools that will ease certain parts of a developer’s workflow, but they will not replace programmers. They push the profession one more rung up the meta ladder—programmers now spend more time directing, verifying, and correcting the automated output.

Asimov knew this intuitively. George Platen’s chosen profession, before he learns the truth, is computer programmer—a far-sighted detail for a story written in 1957, decades before the field was common. The people who write the code that writes code will always be in demand.