About

The throughline

Most systems don’t fit the people who have to use them. Not because anyone designed them badly on purpose, but because the design happened somewhere other than where the work happens. A process gets approved, everyone agrees it’s fine, and then quietly works around it. That gap — between how something is supposed to go and how it actually goes — is where I’ve spent my working life.

The pattern is always the same three moves. Find out where the system doesn’t fit. Build a clearer path through. Then explain it well enough that someone else can walk it without me standing there.

That last part is not an afterthought. A fix nobody can follow is a fix that lasts exactly as long as your attention does.

The professional work

Instructional design, technical training, and increasingly the systems and governance around them — how work gets documented, how decisions get made, and who is accountable when a tool writes the first draft.

I’m most useful on the problems that are underspecified rather than merely hard: the mandate with no template, the course that failed its pilots, the policy that was written for a room the people doing the work were never in. The conduct of operations rebuild is a representative example — inherited mid-development, piloted twice, failed twice, then taken back to analysis and started over when a regulatory order tripled the audience.

I’m moving toward systems-oriented leadership, responsible AI implementation, and project leadership. Not away from the classroom — the classroom is where I learned that the answer is almost never more content.

On AI, specifically

I use these tools daily and I argue for their use. I also think most of the conversation about them is aimed at the wrong question — organizations are focused on what AI can do, and skipping past what it should do. When instructional decisions carry real consequences, that gap is where the risk lives.

My working rule, and the thesis of a conference session I gave for The Learning Guild in August 2026: treat every AI-generated learning asset as a draft that must be traceable, reviewable, and owned by a human. In higher-risk environments, “AI helped me draft it” is not enough. The final product still needs a clear chain of evidence and a person whose name is on it.

I write about this at AI at the Point of Work on Substack, where the recurring subject is the distance between an AI policy and the desk where the work actually happens.

Why the rest of it isn’t a detour

There’s a math channel, a handful of small apps, narration work, and two invented worlds. It looks scattered on a résumé. It isn’t.

Mr. Pledger Stays After School exists because “you’re bad at math” is almost always a misdiagnosis of “nobody showed you the step you actually missed.” That is the same problem as a failed course, at a different scale.

The apps — a meal-plan board, a voice-activated stopwatch, an email server an assistant can use without holding every key — all exist because I needed them, they didn’t exist, and building the missing thing turned out to be faster than continuing to work around its absence.

Even the fiction started this way. Both worlds began as a system I designed and then had to build somewhere for it to exist. Educator, builder, and storyteller aren’t three careers I’m juggling. They’re the same verb pointed at different rooms.

Get in touch

The best way to reach me is email: hello@kwpledger.com

Speaking, training and instructional design work, responsible AI implementation, or a question about something here — all welcome. I answer real messages from real people.