In every organization there's at least one person quietly doing wild things with AI - working faster, thinking bigger, and building their own personal AI stack. In this episode, host Susan Diaz explains how to find those secret power users, support them properly, and turn their experiments into an organizational advantage (without burning them out or making them your unpaid AI help desk).

Episode summary

This solo episode is a field guide to the people already living in your organization's AI future.

Susan starts by painting the familiar picture of the "suspiciously fast" teammate whose work is cleaner, more strategic, and clearly powered by AI - even if no one has formally asked how. She names them for what they are: AI power users who have built quiet personal stacks and are effectively acting as your R&D lab for AI.

From there, she walks through:

• How to spot these people through audits, language, manager input, and usage data.

• Why most organizations ignore or accidentally exploit them.

• A practical three-part framework - Recognition, Resourcing, Routing - to turn power users into supported AI champions instead of secret heroes headed for burnout.

• The equity implications of who gets seen as a "champion" and how to ensure your AI leaders reflect the diversity of your workforce.

She closes with a simple champion blueprint and a piece of homework that any founder, leader, or manager can act on this week.

Key takeaways

You already have AI power users. The question isn't "Do they exist?" It's "Where are they, and what are we doing with them?"

Power users are your unofficial R&D lab. They're not theorising about AI. They're testing it inside real workflows, finding what breaks, and figuring out how to prompt effectively in your specific context.

They are rarely the most technical people. Your best champions are often people closest to the work - sales, customer-facing roles, operations - who are simply determined to figure it out. Not just IT.

If you ignore them, three things happen:

1. They get tired of doing extra work with no support.

2. Their workflows stay trapped in their heads and personal accounts.

3. Your organization misses the chance to scale what's working.

Use the 3 Rs to turn power users into champions:

Recognition - Name the role (AI Champions/Guides), make their contribution visible, and invite them into strategy and training conversations.

Resourcing - Give them real time (10-20% of their week), adjust workload and goals, and reward them properly - ideally with money, training, and access.

Routing - Turn personal hacks into shared assets: playbooks, Looms, internal training, and workflows embedded in L&D or ops.

Connect - don't overload - your champions. Give them a direct line to IT, security, legal, and leadership so they can sanity-check ideas and inform strategy, without becoming the AI police.

Equity matters here. If you only see loud voices and people closest to power, you'll miss quiet experimenters, women, and people of colour who may be building brilliant systems under the radar. Use multiple ways (surveys, nominations, self-identification) to surface a diverse group of champions.

Champions must be guides, not gatekeepers. Their role is to make it easier and safer for others to experiment - not to punish or shut people down.

A simple champion blueprint: identify → invite → define → resource → amplify. Done well, your champions become the bridge between today's experimentation and tomorrow's AI strategy.

Episode highlights

[00:02 ] The "suspiciously fast" colleague and what their behaviour is telling you.

[02:00 ] Personal AI stacks and why Divers "swan dive backwards" into AI without waiting for permission.

[03:37 ] The risk of ignoring power users: burnout, trapped knowledge, and missed scaling opportunities.

[05:03 ] Why power users are effectively your AI research and development lab.

[06:33 ] How to surface power users through better audit questions, open-ended prompts, and usage data.

[07:25 ] Listening for phrases like "I built a system for that" and "I just play with this stuff because I'm a geek."

[08:25 ] Using managers and platform data to spot a small cluster of heavy AI users.

[09:37 ] The danger of quietly turning champions into unpaid AI help desks.

[10:33 ] The 3 Rs: Recognition, Resourcing, and Rooting.

[11:18 ] What real recognition looks like—naming, invitations to strategy, and public acknowledgement.

[12:05 ] Resourcing: giving champions time, adjusting workloads, and updating job descriptions.

[13:14 ] Rooting: creating playbooks, Looms, and embedding workflows into L&D and ops.

[14:29 ] Connecting champions with IT, security, legal, and leadership.

[15:45 ] The equity lens: who gets seen as a champion and who's missing.

[17:16 ] The risk that women and marginalised groups get left behind and automated first.

[18:30 ] Using surveys, nominations, and explicit invitations to diversify your champion group.

[19:07 ] Why champions should be guides, not AI police or gatekeepers.

[19:47 ] The 5-step "champion blueprint": identify, invite, define, resource, amplify.

[22:15 ] Your homework: talk to one secret power user this week and ask how you can make space for their experimentation.

Think of one person in your organization who's already that secret AI power user.

This week, have a conversation that goes beyond "cool, can you do that for everyone?" and into "This is important. How can we make space for you to keep experimenting like this and help others learn from you?"

That's the first step in building your AI champion program - whether or not you call it that yet.

Connect with Susan Diaz on LinkedIn to get a conversation started.

Agile teams move fast. Grab our 10 AI Deep Research Prompts to see how proven frameworks can unlock clarity in hours, not months. Find the prompt pack here.