Building aisa.to — an AI that assesses people's AI skills through conversation.
When we started, we assumed the data would show a simple spectrum: beginners on one end, experts on the other. It doesn't. After 400+ assessments, we've identified 10 distinct personas that keep showing up.
The most interesting one: the Copy-Paster. Uses AI daily, often more than anyone else on their team. But accepts everything at face value. High usage, low skill. These people genuinely believe they're power users because they use ChatGPT 50 times a day — but they never verify output, never adjust their approach, never question what comes back.
Then there's the Sceptic — great critical thinking, barely uses AI. Their caution is actually an asset if they learn to calibrate it rather than avoid the tools entirely.
The tactical insight for founders: "do you use AI?" is a useless hiring question. Two people can both answer "yes, every day" and be completely different personas with completely different skill profiles.
We score across 11 criteria and 5 dimensions, and the persona classification has been one of the most useful outputs for the people taking the assessment. More on the methodology: https://aisa.to/state-of-ai-fluency
Anyone else building assessment or evaluation tools? curious how you handle the "people don't fit neat categories" problem.