Synthesis · May 19, 2026 · Far West Consulting
What Brynjolfsson 2025 means for who you train first
The evidence on who gains from AI assistance is clearer than the usual pilot advice admits, and it argues for training in the opposite order from most rollouts.
For L&D leaders, integration managers, and functional executives planning their next AI training cohort: the evidence on who gains from AI assistance, who doesn’t, and who loses ground is clearer than the usual pilot advice admits.
Brynjolfsson, Li, and Raymond’s 2025 study of a staggered rollout, in the Quarterly Journal of Economics (n=5,172 customer-service workers), found AI raised productivity by 15% on average, with a 30% gain for less skilled and less experienced workers, little effect for the most skilled, and a small decline in the quality of their conversations.7 It maps onto a clean operational rule: AI compresses the gap between the least and most experienced, but it does it by lifting the bottom, not by extending the top.
The conventional pilot playbook says the opposite. Start with the senior team. Build executive buy-in. Cascade from there. The logic feels right because senior people set tone. On productivity, we read the data as pointing the other way.
Three things follow from that.
First, the people with the most to gain are the cohort with the least seniority. New hires, recent rotators, less-experienced practitioners. They’re the cohort where we expect training to show up soonest. Pilots that target this cohort give you the clearest before-and-after work samples, which is the evidence budget holders need to approve the next phase.
Second, senior people need a different curriculum, not the same curriculum applied later. The Brynjolfsson finding doesn’t say “skip the senior team.” It says AI did little for the most skilled workers’ output and slightly lowered the quality of their conversations. The training that works there is judgment-layer training: when to override AI suggestions, how to catch quality regressions in juniors who are using AI well, what kinds of decisions AI should not be near. That’s a different course than the productivity-gain training the junior cohort needs.
Third, the cohort sequence shapes the change story. When this cohort’s agreed number moves at the 30, 60 and 90-day check-ins, the rollout has its own evidence. The senior cohort’s training program is justified by what’s happening one layer below, and that is a defensible story for the board.
This is how we structure the AI training engagements we run. The Discovery shows who’s where on the skill distribution. The curriculum starts with the people who have the most room to grow, and senior judgment-layer training follows, built on what that first group produces.