The "10,000-Step" Rule for AI at Work

A simple & practical habit that ends the “no time for AI” trap

People love a baseline.

In fitness, it’s “10,000 steps a day.”

We know it’s not a perfect scientific threshold, but it works as a behavioral anchor: simple & repeatable (Harvard Health has a great breakdown of the “10,000 steps” story.)

So here’s the question I can’t stop thinking about:

What’s the equivalent of 10,000 steps for AI at work?

Not “nice-to-have.”
Not “when I have time.”
Not “after I finish my real work.”

Non-negotiable. Daily. Baseline.

Meetings are the treadmill we can’t get off

Back-to-back meetings. Decisions made in one call, forgotten in the next.

The hard part isn’t even the meeting.

It’s what comes after.

The invisible second shift that starts when the last meeting ends:

• writing the recap

• reconstructing decisions from memory

• assigning action items

• chasing owners

• sending “quick follow-ups” that are never quick

When calendars are booked back-to-back, recap work spills into after-hours. Into weekends. Into that weird limbo where you are “done working”… but still not done.

Writing meeting minutes is the part that quietly breaks people.

The AI chicken-and-egg loop

This is the line I hear constantly:

“I want to learn AI… I just don’t have time.”

It’s true.

No time means no exploration.
No exploration means no time saved.
No time saved means… still no time.

That loop is exactly why “go play with AI” doesn’t land for most teams.

So the smartest starting point is the one that creates time first.

My baseline: minutes of meeting, every time

If I had to choose one “10,000 steps” AI habit right now, it would be this:

Every important meeting you host gets AI-assisted minutes.

Now you might be thinking, “We can’t do that for every single meeting. There’s confidential stuff discussed. There are privacy concerns.”

I hear you.

I’ll cover that. For now, let’s start with the happy path.

Real minutes:

• what was decided

• what changed

• who owns what

• what happens next

• what’s still open (and what needs a decision)

This is the bridge between “we talked” and “we moved.”

It also pays you back immediately.

No workflow redesign. No big learning curve. Just time returned.

A real time - reclaimer.

The “One Scribe” Rule (this will stay simple even with bots)

Here’s the habit that makes minutes automatic: every meeting has one designated scribe.

Often it’s an AI note-taker.

The point isn’t how many “attendees” show up - some meetings end up with three humans and three bots.

The point is ownership.

Pick one tool as the source of truth, decide where the notes will live, and make one person accountable for sending the recap.

It’s fine if multiple note-takers join.

It’s not fine if nobody owns the minutes.

One meeting. One scribe. One recap.

Make it feel like the Apple Fitness App: close the rings!

Now make the baseline visible.

Every meeting you host = 10 points.

Host 10 meetings, hit 100 points.

Circle closed.

This isn’t about glorifying meetings. It’s about making sure meetings produce movement:

• decisions captured

• owners named

• next steps written down

• clarity shared while it’s still fresh

No more meetings that evaporate into “Wait… what did we decide?”

And no more minutes sitting on your shoulders at 8:30pm.

AI turns the recap from a nightly chore into a default habit.

The excuses I hear (and what to do instead)

People don’t avoid recording because they’re anti-AI. They avoid it because it feels risky, awkward, or exposed.

Here are a few common ones and how to handle them without abandoning the habit.

“I swear too much when we talk about this topic.”

I’ve heard this one. And honestly? It’s human. We need space to vent - for some it could be meetings.

The answer though isn’t “don’t capture the meeting.”

The answer is to separate tone from outcomes.

Try this:

• Keep the capture focused on decisions, action items, and next steps, not verbatim quotes.

• Use the notetaker to produce a short, clean recap: “Here’s what we decided” (not “here’s exactly how we said it”).

• If a segment is going to get spicy or deeply candid, say it out loud: “I’m going to pause recording for 3 minutes.” Then resume when you’re back in decision mode.

“This is sensitive. We can’t record.”

Sometimes that’s legitimate. Not everything should be recorded.

But “sensitive” shouldn’t mean “we leave with nothing.”

Alternatives:

• Capture notes-only (no video), and share a summary that excludes sensitive details.

• Record the parts where decisions are made, pause for the parts where emotions are processed.

• Restrict access so only the organizer (or a small set of people) can see the transcript/recap.

“People won’t speak freely if it’s being recorded.”

That’s a culture signal. Treat it as one.

Two moves help:

• Make the purpose explicit: “We’re capturing this to reduce after-hours recap work and prevent rework. Not to police anyone.”

• Offer an off-ramp: “If you’re not comfortable, we can pause for sensitive segments or do a notes-only version.”

“AI doesn’t summarize well. I write better summaries.”

I hear this a lot. And sometimes it’s true.

The default summary can feel generic. I use M365 Copilot, and I have experienced that.

But that’s not the end of the story. It’s the start of the prompt.

You don’t have to settle for whatever the app spits out on the first try.

Most meeting note tools let you prompt for the minutes format you actually need and once you find a good prompt, you can save it and reuse it every time.

In Teams, for example, Copilot can generate different views of the same meeting depending on what you ask: decisions, action items, open questions, risks, even “what changed since last week.” The summary improves dramatically when you give it a structure to follow.

Try asking for minutes like this:

• “Create minutes with 4 sections: Decisions, Action Items (Owner + Due Date), Open Questions, Risks/Dependencies.”

• “Write a 6-bullet executive recap, then a detailed action list.”

• “List what we agreed to stop doing, start doing, and continue doing.”

• “Pull out only the decisions and owners. Ignore discussion details.”

• “Draft the follow-up email I can send to attendees and anyone who missed it.”

The habit shift is simple: treat the first summary as a draft, then prompt it into your minutes standard.

The privacy checklist (so this stays responsible)

Recording + transcription + AI summaries can be a gift. They can also become a compliance headache if you’re casual about it.

Here are the basics I’d want any team to hold:

1) Consent + transparency
Say it out loud even if the platform already notifies people. A human sentence matters:
“I’m recording the meeting we can send minutes during working hours :)”

2) Access controls
Treat transcripts like internal documents. Decide who should have access, and don’t default to “everyone” if the meeting is sensitive.

3) Don’t record everything
Have a simple mental model:

• routine meetings: okay to capture

• sensitive meetings: capture selectively

• confidential meetings: no notetaker/ recording

The goal isn’t “record all the things.”
It’s “capture what prevents rework.”

4) Store less when you can
If all you need is minutes, keep minutes. You don’t always need the full transcript forever.

5) Keep minutes outcome-focused
Minutes should capture decisions, owners, timelines, risks, and open questions and not play-by-play commentary.

The goal should be to reduce privacy risk and increase usefulness.

Here are some tools to consider if you’re not using them already

Pick the one that matches your meeting platform and is easiest to turn on consistently:

• Microsoft Teams + Copilot (record/transcribe, then summarize outcomes and action items)

• Zoom AI Companion (post-meeting summaries you can share right after)

• Google Meet + Gemini (“take notes for me” inside Meet)

• Otter, Fireflies, Fathom (dedicated notetakers if your org is mixed across platforms)

• Gong (a different angle: communication fitness - talk time, talk-to-listen ratio, and how conversations land)

Choose the tool people will still use on a chaotic Wednesday.

I would love to hear from you:

• What’s your AI baseline - one habit that compounds?

• Where are people trapped in the “no time → no experimentation → no time saved” loop?

• What would change if every meeting had a single designated scribe by default?

• What’s your privacy norm: record everything, record outcomes, or record selectively?

• When work gets busy, do your AI habits disappear… or do they hold?

AI isn’t a one-time adoption.

It’s a maintained way of working.

And minutes are the cleanest place to start.

Stay curious.

AI Lady 💫

About the Author

I’m Priya Tahiliani, and I’ve spent the last 15 years at the intersection of HR and Technology. Most of my career has focused on SAP HCM and SAP SuccessFactors consulting, working with Big Four firms and clients across the globe.

I built and launched my company’s first AI tool by forging a great partnership with IT, and today I continue to work with HR leaders to help shape the future of work with AI.

Beyond work, I serve as Vice President of Public Relations at Toastmasters. I’m also the Founder of the AI Collective – Oakville Chapter in Canada, part of the world’s largest community for AI professionals - a network dedicated to learning and leading responsibly with AI.

And of course, I write the AI Lady newsletter, where I share my experiences, insights, and thoughts about how AI is reshaping our workplaces.