The Method Witness
When output is easy to produce and private stacks are hard to inspect, the useful reference is not a résumé echo. It is a credible observation of how a person works.
Not a claim that a person owns every answer. A statement about what they did when the evidence shifted, the limit appeared, or the work required another look.
01 / a changing signal
A reference used to point backward.
A method witness is someone who can describe how a Stack Person forms, checks, and revises a working judgment without demanding access to that person’s private stack.
A familiar reference says that a person held a role, met deadlines, or was pleasant to work with. Those observations still matter. But they are thin evidence for a world in which a growing share of visible output can be drafted, formatted, and even researched with assistance. A polished document is no longer a reliable window onto the work that made it trustworthy.
App People were often legible because work happened inside a shared interface. The manager could see the queue, the fields, the handoff, and the report. Stack People may work with a more personal arrangement of sources, models, notes, standards, and routines. That can create more capable work. It can also make old forms of evaluation either too shallow or too invasive.
Stack People will need method witnesses: people able to attest to a working practice at its boundary, rather than certify a tool list, a private archive, or a single impressive result.
This is not a proposal to turn every collaborator into a monitor. It is the opposite of a demand for total visibility. A useful witness does not need a password, a prompt history, or access to every discarded thought. They need a chance to observe a consequential piece of work in motion: how the person names the question, handles a limitation, invites correction, and returns when the first answer is not enough.
02 / three useful observations
Witness the points where output stops being enough.
A reference becomes more credible when it describes an observed choice, a boundary, and a revision—not a generic claim about intelligence or tool fluency.
They made the question usable.
Did the person separate a request from the real decision? Did they identify what had to be known, who would use the work, and what could not be assumed? This observation values problem formation over quick retrieval. It tells a future collaborator that the stack begins with a human framing the work.
They kept the limit in view.
Did they distinguish documented material from inference? Did they notice when a source did not travel, a permission was missing, or an answer should remain provisional? A boundary is not a weakness in the signal. It is evidence that the person did not use fluency to conceal uncertainty.
They revised the method.
What happened after feedback, contradiction, or a changed condition? A witness can say whether the person defended a brittle result, updated the work, or changed the process that produced it. This is the observation that separates a one-time performance from a capability that can compound.
03 / resist the false choice
Neither a black box nor a glass box.
Organizations have a legitimate interest in safe inputs, accountable outputs, and work that can be continued. People have a legitimate interest in retaining private notes, personal learning, and methods that they have developed across a life rather than extracted from an employer. The wrong response to that tension is to choose between blind trust and complete capture.
Method witnesses create a third surface. They can describe what occurred at the shared edge: the scope agreed, the sources available, the question raised, the correction made, and the handoff left behind. The organization gains a basis for trust. The person retains the environment that lets their capability keep learning.
This makes references more demanding, but also more useful. “Excellent with AI” is almost empty. “When the source changed, she paused the recommendation, identified the missing permission, and returned with a narrower answer and a clear next decision” is an account of practice.