Public teaching · The project professional
A plausible plan has never been cheaper. A sound one has never been harder to spot.
For project managers, delivery leads and PMOs who now review work they did not draft, and remain answerable for it.
01 · What changed
The cost of a document fell. The cost of being wrong did not.
Project work runs on documents: plans, schedules, risk registers, status reports, business cases. Producing them used to be slow, and the slowness did something useful. It forced the author through the reasoning. Producing them is now fast, and the reasoning is optional.
The artefacts still look right. That is the problem.
02 · The plausibility problem
Fluency is not soundness.
A generated plan is coherent, complete and well-formatted. Whether it is sound depends on things that are not visible in the document: whether the dependencies are real, whether the estimates come from anywhere, whether the risks are the ones that will actually happen on this project with these people.
Reviewing for plausibility is now nearly worthless, because plausibility is the one thing guaranteed. Review has to move upstream, to the specification of what the plan must account for.
03 · Governance and escalation
Escalation is a control, not a courtesy.
In an AI-native workflow, escalation stops being a cultural nicety and becomes a designed control with a threshold. The design questions are concrete: what conditions require escalation, to whom, on what evidence, and what happens if the threshold is met and nobody escalates.
In the Edney Learn Workflow Design Method, escalation is a mandatory control inside Engineer and Govern. It is not a separate stage.
04 · Knowledge continuity
Projects end. The reasoning should not.
Most project knowledge has always been tacit and has always left with the people. AI-assisted work makes that worse in a specific way: the artefact is preserved but the reasoning behind it was never held by anyone in the first place. Recovering and codifying what people actually know is a prerequisite for allocating work between a human and a machine at all.
Recovering and codifying that knowledge is what the ARCHITECT method does. It was originated by Dr Jim Choo and is operated by Edney Learn.
05 · Retained judgment
What a project professional should not hand over.
- Deciding what the plan must account for before it is drafted
- Judging whether an estimate has a basis or merely a number
- Deciding which risks are real on this project, with these people
- Deciding what must be escalated, and doing it
- Standing behind the result
06 · A bounded example
What this looks like on a schedule.
Take a schedule produced in minutes. The AI-native reframing does not ask whether the schedule looks right. It asks: what did we specify it had to satisfy; which dependencies were asserted rather than known; what verification did a human perform, against what; and what threshold would have triggered escalation. The output moves; the specification, verification and escalation do not.
Illustration only. Nothing here assesses the reader or your project.
07 · The Project Manager pilot
The Project Manager pilot.
PilotRunning under controlled conditions to generate evidence. Places, terms and eligibility are set by Edney Learn.
Who it is for. Project professionals carrying accountability for delivery where AI is already in use on plans, estimates, risk registers and reporting.
What it addresses. Work that arrives fluent and cannot be reviewed for plausibility, because plausibility is guaranteed. The pilot builds the specification, verification and escalation controls that replace the review you used to be able to do by reading.
What taking part involves. A defined cohort working on live projects under controlled conditions, run by Edney Learn, who set eligibility, terms and confidentiality.
On evidence. A pilot exists to generate evidence, so there is none yet to quote. No customer outcome, statistic, client name or testimonial is claimed. Do not send live-project material through this website.
Read the wider argument.
This page is one application of a larger claim about how capability is built.