All articles.
Working notes from a practicing ACNP and graduate-program faculty member. Clinical practice, NP education, and the parts of AI showing up in both.
You Can't Stop What's Coming
Slide rules, mainframes, search engines — every tech transition had holdouts right about the tradeoffs, wrong about the timeline. AI is no different.
Read the piece →AI Content Creation Is a Grind. Here's What That Actually Means.
The hype cycle skips the part where you throw away most of what AI produces. Good educational content is iterative and expertise-dependent, not push-button.
Right About the Tradeoffs, Wrong About the Timeline
The holdouts in every technology transition weren't wrong about the costs. They were wrong about the timeline. AI won't wait while faculty debate the tradeoffs.
The Tool Worked. I Didn't.
I built a system to feed my AI the right background automatically, then measured whether it helped. The system got better. I didn't.
The Guideline Is the Step That Gets Skipped
Graduate NP content needs the exact current recommendation, not the version you remember. Verifying it is the first step cut when busy, and nobody catches it.
What Building With AI Costs
A year building products with AI as a primary tool taught me three frustrations nobody warns you about, and results I couldn't have gotten any other way.
Fifteen Minutes, Two Minutes, No Thank You
Twice this week I finished someone else's work in minutes with AI and handed it back. Twice they didn't take it — the reason has nothing to do with the tool.
I Built the AI My Students Use
Two years ago I stopped waiting for someone to build AI tools for my students and built my own. The debate about AI in higher ed asks the wrong question.
The Guideline You Learned Is Not the Guideline That's Current
Practicing APPs work from the guideline version they learned in training. Continuing education tracks hours, not currency — nothing forces an update.
The Lecture I Built On My Way Into the Classroom
I taught a graduate-level lecture on lab-test reasoning the same day I built it. Not faster prep — no prep at all, in the old sense. The tool gave me back my own judgment, ready to teach.
A Benchmark Is a Lab Value, Not an Outcome
New AI models ship every few weeks with a leaderboard screenshot. A former PhD mentor taught me how to judge quality instead, so I built a tool that tests each new model on my own cases against a bar I set before looking.
New pieces, by email.
Clinical education, AI in nursing, and the work of building something real. New articles to your inbox — no noise.
You’re in. First issue coming soon.