AI literacy as a coaching problem: naming the owner

Most teams endorse “AI literacy as a coaching problem” quickly and operationalize it slowly. With AI in the loop, mentorship shifts toward verification, taste, and accountability. Syntax was never the hard part.

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

Write one page while the decision is still warm — context, options, choice, owner, revisit date. Verbal alignment on AI literacy as a coaching problem evaporates under ordinary calendar pressure, and Slack archaeology is a poor substitute for a decision record.

For AI literacy as a coaching problem, 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.

How it actually works

Start with controls that are boring on purpose: written owners, verification, and a path for the next person.

Promotion packets should be assembled from artifacts already in the open: RFCs, incident leads, teaching reviews. If the packet requires secret knowledge, mentorship failed earlier.

Keep a light written trail of growth so progress is not a vibe only the manager remembers. Promotion packets should not require archaeology.

Career conversations should change next week’s work: a stretch ownership, a public design, a safer incident role — not only aspirational titles twelve months out.

Stretch ownership without a safety net is abandonment. Stretch ownership with a named coach, a kill switch, and a debrief is how judgment grows.

Pair on the decision, not only the code. Have the mentee write the tradeoff paragraph first, then implement. Judgment is a writing skill under time pressure as much as it is a coding skill.

Teach the verification step out loud when models draft. Juniors copy what seniors check — or what seniors skip.

When pressure rises, teams drop the unowned practice first. Put AI literacy as a coaching problem in the path of work — templates, checklists, review norms — or admit it was optional applause.

Where this shows up

AI-assisted workflows arrive without coaching norms. People paste secrets into tools, accept insecure snippets, and skip tests because the model “looked sure.” Literacy clinics on one failure mode per month beat a single all-hands about AI.

A mentee ships a migration with a written plan and a debrief for the team. That single loop teaches more than six months of vague “visibility” talk.

A senior is excellent individually and absent as a teacher. Promotion criteria still overweight personal output. Leadership then wonders why the bench is thin. Change the scoreboard: teaching artifacts count.

Tradeoffs

With AI drafting, the tradeoff is review load. Explicit verification notes reduce merge regret more than asking people to “be careful.”

The tradeoff in public sponsorship is risk to your own reputation. Take it. That is part of the job. Private cheerleading costs you nothing and buys them little.

The tradeoff in not stealing the work is short-term speed. You will be slower this afternoon so the team is faster next month. Leaders who cannot tolerate that tradeoff become permanent critical path.

Seniors need deliberate practice teaching. If the scoreboard only rewards personal velocity, judgment does not scale past one hero.

Resist the urge to expand scope into neighboring slogans. If a control does not make AI literacy as a coaching problem more operable for the next person, leave it for another note. Dilution is how coherent essays become stitched scrapbooks.

Failure modes

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

• AI tools introduced without coaching on verification, data boundaries, or accountability.
• Mentorship as calendar filler: 1:1s that never change assignments or ownership.
• Feedback that arrives only at review time, when grades feel political instead of formative.
• Stealing the work under the banner of helping — the mentee never practices judgment under load.

What to do this month

• Every PR review: at least one comment that teaches a reusable pattern.
• Every month: one public sponsorship act — authorship, presentation, or incident lead.
• Every 1:1: one concrete coaching ask, one ownership shift, one written follow-up.

Critique privately and specifically. Vague encouragement teaches nothing. Vague criticism teaches fear. Specific critique tied to a reusable pattern teaches judgment.

When someone freezes asking for help, treat it as a systems smell. Latency to ask questions predicts incident severity. Psychological safety is an engineering control.

The psychology of asking for help is shaped by what got rewarded last time someone asked. If asking is punished as weakness, AI tools will hide mistakes instead of surfacing uncertainty.

Sponsor publicly. Put someone’s name on the design review invite, the RFC author line, the customer-facing postmortem. Private belief in someone does not compound; public sponsorship does.

Seniors need deliberate practice teaching: lead a review huddle, write the onboarding doc, own feedback quality. If you only reward personal velocity, teaching becomes unpaid overtime.

Pairing as coaching is not surveillance. Time-box it, name the learning goal, and leave an artifact. Surveillance pairing produces compliance; coaching pairing produces judgment.

Close

The measure of mentorship is whether the next hard problem can be owned by someone else — with you still available, not required.

If someone new cannot explain AI literacy as a coaching problem from your artifacts in fifteen minutes, the practice is still private.