A talk · NACWPI / CMS
Playing Out of Position
The Adaptive Expert
“…or other duties as needed by the unit.”the line in every academic job description
How a performer keeps getting drafted into roles they never trained for — and uses AI to suit up.
Jason M. Johnston, D.M.A.Associate Professor of Music · University of Idaho
The thesis
It’s adaptive expertise, not faking it
The self-doubt is the hook — then the flip. A routine expert masters the thing they trained for; an adaptive expert extends real skill into genuinely new territory. The apparent lack of credentials, reframed, is the most transferable thing I have to teach.
“I am not an expert at any of the things I’ve brought to the table. I am a performance-based researcher. I’m not a musicology PhD; I have no conducting credentials. I do these things because I was asked to, and I figured it out.”
The line that keeps it honest: my real, credentialed expertise — the ear, the score, research through performance — is what lets me direct and audit AI in the domains I’m still learning. I extend; I do not defer.
The method is constant — AI is just a new consultant
- 1Say yes
- 2Consult expertise
- 3Audit yourself honestly against it
- 4Iterate by doing
AI is another colleague you consult and audit — fast, always-available. Not an oracle.
The arc — click a case to explore
Four times I was asked, and figured it out
The 2018 band is the precedent — I’ve always said yes. But the real story is the recent cluster: starting in 2025, a dramatic shift from a comfortable, forward-moving researcher to a teacher treading water — handed course after course I was never trained for. AI is how I’ve stayed afloat.
The concert band
Asked to conduct as a "short-term bandaid." Did it so well they stopped searching — I became it. No credential; consulted colleagues around the country; learned by doing. AI absent by nature.
The baseline caseMusic Since 1900 (MUSH 4190)
A 4000-level music-history course — "I could not be more meekly qualified." The music is mine; the scholarly dissemination was the gap. I built the architecture; AI implemented it.
The machine I builtLHSOM & university assessment
The same deluge at institutional scale — I built the school's entire assessment system, NASM-aligned, across all eight degrees. Never trained for it.
Assessment at scaleGrad Research & Bibliography (MUSX 5110)
Back to the surface — "bringing it into the 21st century with a significant AI module." The recursion: AI is scaffold and subject — now I teach the loop.
The loop, taughtSaid yes (2018) → Thrown in the deep end (2025) → Afloat with AI → Teaching the loop forward (2026)