Entry-level user experience work is disappearing from the desks of junior designers and researchers, and with it the daily practice that once turned beginners into skilled practitioners. Tasks like tagging interview transcripts, drafting first-pass wireframes, and marking up redline sheets have been quietly absorbed by artificial intelligence tools that now return sessions tagged, themed, and clipped before anyone opens the recording. The output is often good enough, and nobody argued when the work stopped being assigned.
What went unrecorded was what those tasks taught. Tagging transcripts is how a young researcher learns that people say «it's fine» when they actually mean they gave up, a nuance automated tools do not catch. Marking redlines is how a new designer learns which spacing decisions the design system made and which remain open questions. The work was never valuable as a deliverable, but it was invaluable as supervised practice, where being wrong cost nothing and the lesson stayed with the person who did it.
The employment numbers already reflect the shift. Research by Erik Brynjolfsson, Bharat Chandar, and Ruyu Chen, drawing on payroll records through June 2026, found that employment for workers aged 22 to 25 in the most AI-exposed occupations sits 19 percent below where it would be had it kept pace with less-exposed peers. No comparable gap exists for experienced workers. The adjustment is happening through hiring rather than layoffs, meaning the entry ramp itself is narrowing.
The researchers propose a mechanism that design leaders should consider carefully: AI substitutes for codified knowledge and complements the tacit knowledge that is built only by doing. If junior designers and researchers are not doing the doing, they are not building the tacit skills the field depends on. A 2026 report from Maze found that 69 percent of research professionals now use AI in at least some projects, up 19 points in a year, concentrated in transcription, synthesis, and drafting questions, precisely where juniors used to start.
Ethan Mollick has argued in «Choosing to Stay Human» that the defaults for what work gets handed to AI are being set right now, mostly without planning, and will be hard to reverse once a generation of workers builds habits around them. The recommendation for design teams is to inventory the tasks automation has removed over the past 18 months and, alongside each one, record what that task taught the person doing it. On most teams the list includes transcript tagging and synthesis, first-pass wireframes, screener drafts, competitive audits, redline sheets, and meeting notes. Each was cheap to automate and expensive to lose as training.
The inventory should be run with senior staff in the room, because they can recall the specific lessons embedded in routine work. A screener draft, for example, was where many researchers learned to write a question that does not leak its own answer. Without those stories, the list looks like a set of chores and teams may conclude nothing was lost. A task can be worthless as output and foundational as training, and that pairing is exactly what automation dissolves.
There is a parallel in surgical training. Matt Beane spent two years observing robotic and open operations for his research on robotic surgical skill. In open surgery, the resident's hands were required. On the robot, the attending physician can work alone while the resident watches a screen, and sometimes the robot performs skills the surgeon should be practicing. A 2025 national survey of recent U.S. general surgery graduates found only 37 percent reported high autonomy in robotic cases, compared with 89 percent in open ones. The mechanism transfers to design even without a credentialing gate: when the expert can finish without the novice's hands, the novice stops learning.
The proposed remedy is to assign juniors real decisions with guardrails. Let them choose which direction goes to review or which participants come out of the recruit. The senior asks why, evaluates, and revises while the junior takes notes and watches the screen. Everyone stays busy, but only one person is learning. Building that practice back deliberately is how the field keeps its human element and builds the bench it will need for 2030.