AI in the workflow, not on the slide deck

The useful version of “AI in the workflow, not on the slide deck” is narrower than the slogan: Fluent wrongness is the new failure mode. Confidence in prose is not evidence.

What follows stays on that claim: mechanisms, concrete scenes, and the failure modes that appear when the claim is ignored.

When pressure rises, teams drop the unowned practice first. Put AI in the workflow, not on the slide deck in the path of work — templates, checklists, review norms — or admit it was optional applause.

Shared harnesses beat private prompt folklore. If the “right way” lives in one chat history, you do not have a team practice — you have a bus factor dressed as productivity.

A scene from delivery

Someone pastes a production schema into a consumer model tool to “move faster.” Leadership had no published data boundary. That is not an individual ethics failure first — it is a missing control.

That scene is the thesis under calendar pressure. The response is not another status meeting — it is a control.

Mechanisms

Shared harnesses beat private prompt folklore. If the only way to do it “right” lives in one engineer’s chat history, you do not have a practice.

Require a verification trail on assisted work: what the model changed, what a human checked, and where that note lives — PR, ticket, or decision record. Authorship is not ownership of production risk.

Update review norms when copilots arrive. If you do not, you will approve fluent bugs faster. Slow the merge until verification is explicit.

Prompting is not process. Harnesses are: allow-lists, templates, evals, and review norms that survive a teammate leaving.

Treat “the model said so” as a non-reason in design review. Ask for the invariant, the test, or the threat model instead.

Measure assistance by outcomes you already care about — change-fail rate, review cycle time, incident escape — not vanity “lines generated” or “acceptance rate.”

Write one page while the decision is still warm — context, options, choice, owner, revisit date. Verbal alignment on AI in the workflow, not on the slide deck evaporates under ordinary calendar pressure, and Slack archaeology is a poor substitute for a decision record.

Another texture

A team bans AI briefly after a bad merge, then reintroduces it without new controls. The ban was theater; the missing piece was verification and data boundaries.

A literacy clinic walks through one authz bug introduced by an accepted suggestion. The room learns more than from a generic AI town hall.

Tradeoffs

Tempo gains are real on boilerplate and real-negative on subtle invariants. Teach the difference explicitly.

Verification notes add friction to the PR. They remove friction from the incident. Budget for the former.

Restricting tools feels anti-innovation until the first compliance event. Policy is what lets you expand autonomy later.

Keep the unit of progress small enough to finish under a full calendar: one owner clarification, one verification signal, one reversible control. Grand programs without weekly evidence become status machines.

For AI in the workflow, not on the slide deck, the inheritance test is blunt: after a week, can someone outside the original room explain what changed, who owns it, and how you will know if it breaks? If the answer depends on hallway memory, you still have a story — not a practice.

Failure modes

Each of these is a missing control, not a personality problem:

• Leaders celebrating velocity while review quality quietly collapses.
• Private prompt folklore instead of shared harnesses and policies.
• No exit criteria for tools — the vendor becomes the process.
• Treating model output as authority because it sounds senior.

What to do this month

• Track escaped defects that originated in assisted drafts and coach from that set.
• Add a “human verified” checklist to PRs that used model drafts.
• Publish allowed tools and prohibited data classes in the engineering handbook.

Shipping with models without losing craft requires the same standards you already had for human drafts, plus an explicit verification step because fluency hides gaps.

Teach distrust of fluent wrongness. Models sound senior. Tests, invariants, and threat models still decide truth.

Evaluating AI assistance without vanity metrics means comparing escaped defects, onboarding time, and review quality — not counting accepted suggestions like arcade tickets.

Define allowed tools and data boundaries. Copilots without policy become shadow IT with autocomplete. Secrets, customer data, and authz logic need explicit rules.

Coach AI literacy as judgment: when to draft with a model, when to refuse, when to escalate. Literacy is a coaching problem, not a one-time workshop.

Publish what must never go into a model tool: secrets, unrestricted customer data, unresolved legal holds. Ambiguity here becomes shadow IT with a chat UI.

Keeping humans accountable when models draft code means the merge still carries a name and a verification note. Anonymized autocomplete is not an accountability model.

Where copilots help: boilerplate, test scaffolding, exploratory refactors with strong suites. Where they hide debt: security boundaries, concurrency, and domain invariants nobody wrote down.

AI changes tempo; leadership still owns outcomes. If escaped defects rise while suggestion counts rise, you optimized the wrong dashboard.

A strong IC becomes a bottleneck because only they know the “good prompts.” Codify the harness: templates, allow-listed tools, eval snippets, and examples of refusals.

Close

Tempo is a gift. Accountability is still the job. Keep both in the same sentence.

If someone new cannot explain AI in the workflow, not on the slide deck from your artifacts in fifteen minutes, the practice is still private.