Before we assess people’s AI skills, we should ask how they feel
Every company rolling out AI tools runs into the same quiet problem. The tool works, the training exists, and still most people don’t use it. On paper the change is human-centred. In practice, many employees feel overwhelmed.
In my report on the NECP Framework (Naming Emotions Collaborative Prompting), I argue that part of the problem is how we measure learning. Skills assessments are usually built to compare. They rank people, show how many colleagues did better, and treat one bad Monday as a fair picture of someone’s ability. People feel judged, so they hold back, answer defensively, or avoid the survey altogether. Then we make decisions from that data.
NECP starts with one simple step: name the emotion first. Before a short skills check, a chatbot asks how you feel. You pick a broad emotion, then a more precise one. Naming a feeling doesn’t make it disappear, but it does make it less likely to quietly drive your answers. What a manager sees as “resistance to change” is often comparison anxiety. What looks like conflict or burnout is often an unmet need that nobody has said out loud.
Two more ideas complete the framework:
Self-benchmarking. You are only compared with your own earlier results, never with your colleagues. Progress becomes something you see, not something you’re ranked on. Complementary pairing. Some people learn fast and then stall. Others start slowly and build steadily. NECP looks at these learning curves and pairs people for recurring peer sessions. The fast learner gives momentum and the steady learner gives persistence.
The report illustrates this with Theodor, a lawyer who finds a new AI data catalog overwhelming. Week after week, the bot’s check-ins make it easier for him to admit how he really feels. Meanwhile he is paired with Maria, a colleague who already uses the tool well. Three months later, Theodor is more confident than he expected to be. Maria, by explaining the tool, understands it better than before.
As AI takes over more routine work, the skills that matter most are judgment, imagination and empathy. It would be odd to build those skills with assessment methods that make people anxious.