Beyond Users: Rethinking How We Refer to People in the Age of AI

It is a mainframe-era leftover that outlived the machines it was built for.

"User" entered tech in the 1950s to describe the trained staff who operated room-sized mainframe computers, and it stuck through every shift since - personal computers, the web, mobile, and now AI - long after the world it described disappeared.

Writing in MIT Technology Review in 2024, Taylor Majewski traced the term back to the mainframe age, when a "user" was one of a handful of employees cleared to operate an expensive machine. As computing spread to desks in the 1970s and to everyone in the 1990s, the word came along for the ride. "User account," "user ID," and "user profile" hardened into plumbing nobody questioned. The label survived because it was convenient, not because it was accurate.

Don Norman helped cement it. He joined Apple in the early 1990s with the title "user experience architect," which put the word at the center of an entire design discipline. Then he turned on it. In a 2006 essay for the ACM journal Interactions, titled "Words Matter. Talk About People: Not Customers, Not Consumers, Not Users," Norman made the case that the term flattens people into a function. His line has stuck with a lot of us in this field: psychologists depersonalize the people they study by calling them "subjects," and technologists do the same by calling them "users." Both, he wrote, pull you away from the actual job, which is helping people.

Others floated replacements. In her 2011 book Inventing the Medium, Georgia Tech professor Janet Murray argued for "interactor," a word that credits people with creating and participating rather than just consuming. Facebook shifted toward "people" around 2014. Per Majewski's reporting, Jack Dorsey pushed Square to swap "user" for "customer," on the logic that "customer" sets a higher bar for the service you owe someone. None of it fully took. "User" is still the reflex word, which is exactly why it is worth stopping on.

Because the label you use for someone quietly sets how much you owe them.

Calling people "users" does damage because it swaps a whole person for a function, and a function is easy to count, easy to optimize against, and easy to stop seeing as a person you are responsible to.

Norman's argument was not sentimental. It was about consequences. When early technologists treated the user as, in his words, "simply another component" of a system, they built systems that served the system. That framing scales badly. It is a short hop from "the user" to a dashboard number, and a dashboard number does not have a bad Monday, a payroll deadline, or a team depending on the network staying up.

The word carries other freight too. "User" is also the language of dependency and addiction, and given how much modern software is engineered to keep people scrolling, that overlap is not entirely an accident. And the vagueness of the term greased the "growth at all costs" era, where companies chased user counts instead of asking whether any actual person was better off. Count users, and you optimize for more users. Serve people, and you optimize for something harder to fake.

None of this means "user" is a slur or that everyone who says it is careless. Instagram's own chief has used "people" and "users" in the same breath, and the industry runs on the word. The point is narrower and more useful: the vocabulary a company reaches for by default tells you something real about where its attention goes.

Because the machines got the warm names and the humans kept the cold one.

AI sharpened the question because generative tools arrived wrapped in humanizing titles - "assistant," "copilot," "collaborator" - while the people using them stayed "users," an inversion that quietly hints at which side the industry is centering.

Look at how the new tools were introduced. They are not "the automation." They are your copilot, your assistant, your collaborator, language chosen to signal partnership. Majewski flags this directly: the software gets promoted to teammate while the human keeps a decades-old functional tag. Some researchers have even started describing language models themselves as a kind of user the software must serve. Set that beside a person still filed as "user account," and the hierarchy feels upside down.

I am not arguing AI is plotting anything. It is not. But the words we pick around a technology shape the relationship we build with it, and right now the words are drifting toward treating the tool as the important party. For a business owner, that is worth noticing, because you are about to hand more of your operation to these systems. The question underneath the vocabulary is a practical one: is your technology, and the provider who runs it, organized around you, or around itself?

Norman put the stakes plainly: we change the world, and the world changes us back, so we had better be careful how we change it. That is not soft advice. It is the whole reason a business should care which words its technology partner lives by, especially as AI moves from novelty to the thing running the back office.

CinchOps is a managed IT services provider based in Katy, Texas, serving small and mid-sized businesses across the Houston metro area.

CinchOps specializes in cybersecurity, network security, managed IT support, VoIP, and SD-WAN for businesses with 10-200 employees. The language in this post is not marketing dressing - it describes how the work actually runs. We built the company around the idea that the people we serve are partners, not "users," and that shows up in the ordinary details:

• We learn your business first. Instead of "user onboarding," we start by understanding what your company does, who is on the team, and what breaks your day when the technology fails.
• We measure outcomes, not activity. The number that matters is whether your business runs better, not how many tickets closed or how many seats logged in.
• You get a team that knows you. Support is not an anonymous queue. It is people who recognize your name, your setup, and your history.
• We keep you in charge of the AI, not the other way around. As these tools move into your operation, we make sure they serve the people doing the work.

CinchOps serves businesses across Houston, Katy, and Sugar Land, with industry experience in law firms, CPA firms, and construction - the kind of companies that get treated like account numbers by big providers and deserve better.

You are not a user. You run a business, and you deserve an IT provider who treats you like it. If that is the relationship you want, talk to CinchOps and see what it feels like to be a partner instead of a number.