NN/g: UX Is Becoming Experience Design as Agents Replace Users

Two articles published by the Nielsen Norman Group on 30 October 2026 make an argument the field has been circling for years: experience design is the better name for what practitioners now do, because the person using your product is increasingly not a person.

The first piece, Experience Design: Dropping the “User” from “User Experience”, lays out the naming case. The second, The Role of UX in Agentic Software Factories, lays out the structural one. Read together, they say something practical about how you should describe your work in 2026 — and what you should refuse to give up.

The agent is now the user

NN/g’s argument starts with a shift that has nothing to do with chatbot widgets. Don Norman coined “user experience” in the 1990s to widen a field that was stuck on “usability” and “user interface.” For decades one assumption held: the human is the user, interacting directly through channels the company designed.

AI breaks that. A person can ask an assistant to book a flight, cancel a subscription, or file a complaint. The agent becomes the immediate user of the digital product while acting on the human’s behalf — and it may operate entirely outside the touchpoints, journeys and service blueprints your team mapped.

That is the whole naming problem in one sentence. If the direct user is software, then “user experience” describes only half the interaction. NN/g’s position is that UX is not disappearing; the field is expanding in several directions at once, and experience design is the name that covers the wider object.

What the agentic factory leaves out

The second article is the harder one to hear. In August 2026, Anthropic published its vision of an AI-native software development lifecycle, following similar — though less prescriptive — publications from Google in May 2026 and OpenAI in November 2025. NN/g’s read: all three sideline UX, reducing it to behind-the-scenes context that guides AI generation.

The logic each company makes is straightforward. AI agents make writing code fast and cheap, so every other stage must adapt to keep pace. Google’s summary of the shared argument: the shift is “from writing code to expressing intent, and trusting intelligent systems to translate that intent into working software.”

NN/g’s objection is that context is only one third of the value. UX also discovers what users need, strengthens the team’s shared judgment about what to build, and evaluates what gets built before and after launch. An organisation that adopts the model as described risks losing the very user insight its AI context depends on.

Three roles that do not reduce to a prompt

Here is the part worth acting on. NN/g frames an AI-native lifecycle as needing all three UX roles, not one:

• Discovery — finding out what users actually need, which no context library supplies on its own.
• Shaping — strengthening the team’s shared judgment about what is worth building.
• Evaluation — checking what gets built, before and after release.

If your current job description reads mostly as “write specs and prompts for the build stage,” you are sitting in the role these articles say is being squeezed. The counter-move is not to resist agents. It is to make discovery and evaluation visibly indispensable — the two things a context file cannot do for you.

What to call yourself in 2026

NN/g is candid that it has been here before. The group changed its own job titles from “user-experience specialist” to “experience specialist” and, by its own admission, the original explanation did not convince anyone — including its own author. The new argument is different: the object of design has genuinely widened.

For a freelancer or product designer updating a portfolio or LinkedIn headline, the practical question is whether to switch terms now. A useful rule: use “experience design” when your work spans agents, service touchpoints and post-launch evaluation; keep “UX” when the role is scoped to a product surface people click. Recruiters still search for UX. Clients search for outcomes. You can hold both.

Where this leaves your toolkit

None of this changes what you open on Monday. It changes what you claim you are doing when you open it. If your work is heading toward agentic workflows, the tools that matter are the ones that let you keep context structured and reviewable rather than buried in chat history — because the discovery and evaluation work NN/g describes has to live somewhere your team can see it.

The honest caveat: NN/g is arguing a position, not reporting a study. Neither article carries survey data. What they carry is a defensible framing — and in a week where two pieces landed on the same day, it is the framing most likely to show up in your next performance review or client pitch.

Sources

• Experience Design: Dropping the “User” from “User Experience” — NN/g
• The Role of UX in Agentic Software Factories — NN/g