On most small teams right now, something uncomfortable is happening.

One or two people have quietly become AI power users. They're moving faster, producing more, handling work that used to take twice as long. They may not even be advertising it.

Meanwhile, the rest of the team is using AI occasionally, sporadically, or not at all.

That gap — between your most and least AI-fluent employees — is growing. And it's creating problems you might not see yet.

What the Gap Looks Like in Practice

The best AI user on your team is probably:

  • Drafting emails, proposals, and summaries in a fraction of the time
  • Doing pre-call research in 8 minutes instead of 45
  • Using AI to synthesize reports they'd have spent an hour reading manually
  • Building and iterating on prompt templates that compound over time

The least AI-fluent person on your team is probably:

  • Doing the same tasks the same way they've always done them
  • Occasionally trying AI, getting mediocre output, and going back to manual
  • Spending extra time on work that's become a commodity for their more fluent colleagues

This isn't about intelligence or effort. It's about exposure, practice, and the feedback loops that build skill. The power user got enough early wins to keep going. Everyone else hit early friction and didn't have a reason to push through.

Why This Matters More Than Most Owners Realize

The gap creates four specific problems.

1. Silent inequity in workload.

If one person can produce in 4 hours what takes another 8, you either end up overloading the fast person or underutilizing the slow one. Neither is fair, and neither scales. At some point, the gap in output becomes visible and creates tension.

2. Knowledge stays siloed.

The power user has built a prompt library, a set of workflows, and a mental model of where AI works and where it doesn't. That knowledge is sitting in their head (or their personal docs) rather than being shared across the team. If they leave, it walks out the door with them.

3. Your team's productivity ceiling drops.

As AI tools get better and more integrated into daily work, the gap between the fluent and the non-fluent will only widen. The teams that close this gap now are building a durable advantage. The teams that don't are accumulating a skills debt that gets harder to pay off over time.

4. It's demoralizing for the people who are behind.

Nobody enjoys feeling like they're falling behind. Employees who sense the gap but don't know how to close it often disengage from AI entirely — which widens the gap further. The learning curve stops feeling like an opportunity and starts feeling like a verdict.

How to See the Gap Clearly

Before you can close the gap, you need to make it visible. Most owners don't have good data on this — they have a vague sense that some people are using AI and some aren't, but nothing specific.

A few ways to get clearer:

Ask directly. A simple 3-question check-in works: Which AI tools are you using regularly? What's one task where AI has saved you meaningful time? What's one task you've tried AI for and abandoned?

The answers reveal a lot. People who are genuinely using AI have specific answers. People who aren't tend to answer vaguely or describe one or two isolated experiments.

Watch for the outputs. If one person consistently produces polished first drafts while another always submits rough work, that's often an AI fluency signal — not a quality signal. The underlying skill level may be equal; the tool fluency isn't.

Look at who asks for help with AI. In most teams, the people who are actively building AI skills ask questions and share discoveries. The people who've disengaged from AI go quiet on the topic. That silence is data.

How to Close the Gap (Without a Training Program)

You don't need a formal program. You need three things.

1. Surface the power user's knowledge.

Ask your best AI user to share what's working. Not in a big presentation — just a 15-minute informal session showing two or three workflows. Or a short document: "Here are the three prompts saving me the most time right now."

This does two things: it transfers knowledge, and it normalizes AI use by showing that a peer (not an expert from outside) has figured this out.

2. Give the struggling users a specific starting point.

Generic advice ("just try ChatGPT!") doesn't work. A specific starting point does: "For your role, try this for this one task this week. Here's the prompt. Here's what to look for."

The barrier isn't willingness. It's not knowing where to begin with work that actually matters to them.

3. Create a shared space for wins and failures.

A Slack channel, a shared doc, a 5-minute segment in the weekly standup — any low-friction channel where people can share what they tried and what happened. Include failures. Normalize the learning curve. Make it safe to experiment and report back.

The power user didn't get there by having a talent for AI. They got there by taking more swings. You're just creating conditions where everyone can take more swings.

What Closing the Gap Looks Like at 6 Months

When you invest consistently in closing the fluency gap across your team, the shift is measurable.

The outlier power user is still ahead — they always will be, because they'll keep learning faster. But the floor lifts. The least fluent person goes from "barely using AI" to "using it confidently for 3-4 workflows." The gap between best and worst narrows. Knowledge circulates instead of staying siloed.

The business outcome: your whole team's output capacity increases, not just one person's. That's the compounding return on learning culture.

OpenSkills AI gives you visibility into the fluency gap across your team — skill tracking that shows who's progressing, who's plateaued, and who needs a more specific starting point. Role-based learning paths for everyone, not just the people who figure it out on their own.

See how it works for your team or start your free trial.

For the foundation on building this kind of team culture, read what a learning culture looks like at a 12-person company.