Human Adoption & Change Absorption

What this skill is for

AI compresses the production side of work dramatically. It does not compress the human side: judgment, context-loading, alignment, trust-building, and change absorption. This skill exists to close that gap — ensuring that fast AI output translates into real human action, not stalled inaction.

The core problem: when AI produces high-quality output quickly, it looks like things are further along than they are. A polished doc isn't shared understanding. A completed plan isn't a committed team. The artefact exists; the humans haven't caught up to it yet.

The two clocks rule

Production speed and integration speed are different clocks. They must be deliberately decoupled.

Just because something can be produced in a day doesn't mean it should land, get reviewed, and move to execution in a day. When these clocks aren't managed separately, the result is outputs that are ready but unacted on — not because people disagree, but because they haven't absorbed.

In practice: build pacing into every initiative. Name when the output will land separately from when action is expected.

The absorption gap

People who use AI themselves are keeping up because they're processing with AI in real time — it's part of their thinking loop. People who receive finished outputs are consumers, not participants. That's a fundamentally different cognitive experience.

The embracers feel momentum. The others feel a pile of things they haven't quite caught up to yet. And when you're behind, the path of least resistance is to not move at all.

The result is stalling, not resistance. Don't treat stalling as disagreement — treat it as incomplete absorption.

The legitimacy problem

When AI-generated work lands without human fingerprints, the first response is often "this doesn't reflect reality." That is not a comprehension problem. It is a legitimacy and ownership rupture.

Signs of a legitimacy problem:

"This was dropped on us"
"This doesn't reflect how things actually work"
Polite nodding followed by no action
Questions that reopen decisions already made in the document

The fix is not explanation. It is restoration.

Before explaining content, acknowledge that the artefact is a starting point, not a verdict. Transfer authority back: their reality is the input that makes it useful, not the thing it needs to override.

A legitimacy-restoring message has three moves:

Name the AI involvement directly — remove the uncanny feeling of a document that appeared from nowhere
Position the artefact as structural scaffolding, not truth
Give people a role as critics and contributors, not approvers

The handoff layer standard

Any output Tiffany produces for others to act on should include a handoff layer. This is a short wrapper — not the full document — that answers four questions:

What is this? One sentence on what it is and why it exists.
What's already decided? What doesn't need their input.
What still needs you? The specific contribution being asked for — named per person where possible.
One action. What they need to do, by when.

The full depth stays available but is not the required entry point. People should be able to act from the wrapper alone.

Apply this automatically to any artefact, Notion doc, process map, or brief Tiffany will share with the team.

Session design: builds vs briefings

Briefings produce receivers. Builds produce owners.

A briefing: Present the artefact, walk through it, invite questions, assign actions. A build: Get their answer first, then use the artefact to pressure-test and fill gaps.

The difference in ownership is significant. People who constructed something — even partially — feel accountable for it. People who received it do not.

Principles for designing a build session:

Open with a question that gets their view before showing anything: "What's your instinct on how this should work?"
Use the artefact to interrogate their answer, not replace it
Close by having them write the output in their own words
Reserve at least one genuine input point per person — something where their answer could actually change the outcome
Never use the full document as the room's centrepiece

For the 60-min working session format:

Pre-read lands 48hrs before: context (2–3 sentences), the one question the session needs to answer, what each person specifically needs to bring
Session opens with their answers, not a presentation
Session closes with their version of the process, not adoption of yours

Output design triggers

When producing anything for Tiffany to share with the team, apply this checklist automatically:

Does this have a handoff layer (what/decided/needed/action)?
Is the ask isolated and unambiguous — one thing, not buried in context?
Is there a genuine input point, or is this approval-only?
Is the cognitive load of engaging with this small enough to actually happen?
Are the two clocks managed — when does this land vs when is action expected?

If producing a session plan, apply the build vs briefing check:

Does the session start with their answers or our artefact?
Does each person have a named role, not just attendance?
Does the session produce something in their words?

Recovery: when adoption has already stalled

When work has already landed and people are lost or disengaged, the recovery sequence is:

Name it — acknowledge directly that the artefact landed without enough context or involvement. Don't defend it.
Restore legitimacy — position the artefact as scaffolding, not verdict. Their reality is the thing that makes it useful.
Give them a role as critics — "come ready to poke holes, not to approve" is more activating than "please review and provide feedback"
Shrink the immediate ask — the next step should be tiny. One question, one reaction, one decision. Not a full review.
Then rebuild toward a proper build session — don't skip straight to execution from recovery

Principles summary

Situation	Default move	Better move
Sharing an AI-generated artefact	Send the doc	Send the handoff layer; doc is available
Running a working session	Walk through the plan	Open with their answer; use the plan to test it
Adoption has stalled	Explain more	Restore legitimacy first
Work is ready to move	Move immediately	Decouple production readiness from human readiness
Team isn't acting	Assume resistance	Diagnose absorption gap first

human-adoption