THIS MONTH'S FIELD SIGNAL

Activity Without Accumulation 

Something keeps coming up in my conversations with leaders lately — not as a complaint, exactly, and not as a crisis. More like a quiet unease that surfaces when I ask a simple question: what has your organization actually learned from its AI work so far?

The pause that follows is usually the answer.

Most organizations have a lot of AI activity on the books. Pilots launched, licenses purchased, training completed, use cases documented. By most measures, things are moving. But when you ask what carried forward — what the next team will know that the last one didn't, and what the organization would do differently if it started over today — the answer is often murky.

The metrics track effort, not accumulation. You can count what was tried. You can't easily describe what stuck.

What I'm hearing sounds like this: individuals are figuring out AI largely on their own. Someone finds a prompt or a workflow that works well for a recurring task; nobody else finds out. A pilot produces real insight about where AI helps and where it doesn't; that insight lives in the memory of the team that ran it, and the next team starts close to zero. The organization accumulates experience without accumulating knowledge.

This pattern is showing up now because most organizations have moved past the access phase. Tools are deployed. The initial wave of curiosity has happened. But the infrastructure for converting individual discovery into shared practice was never built alongside it.

That infrastructure doesn't have to be complicated. It isn't a formal knowledge management system or a lessons-learned database that nobody uses. It's something simpler and harder: a norm that says sharing what you've figured out — including what didn't work — is a contribution worth making. And a structure, however lightweight, that makes that sharing routine rather than exceptional.

Without it, every person in the organization stays on a solo journey. The activity is real. The accumulation isn't.

FROM THE FIELD

A Real-World Example

A senior operations leader at a mid-size professional services firm is several months into an AI pilot with a small subset of her organization. The tools are live, and people are using them to summarize documents, handle repetitive tasks, and move a little faster on things that used to take longer. A quick internal survey suggests they're positive about it.

But when she describes the state of things, what emerges isn't momentum. It's a collection of individuals figuring things out alone, with no shared framework for what they're doing or why. People have questions they don't know where to bring. Enthusiasm exists, but it's scattered. No one is quite sure what anyone else is trying, what's working, or where the real opportunities are.

She can see use cases where AI could create real organizational efficiency — not just individual time savings — but there's no structure for surfacing those, testing them, or turning individual discovery into shared practice.

What's missing isn't enthusiasm, and it isn't capability. It's a common language and a place for the learning to land. Right now, everyone is on a solo journey. The activity is real. The accumulation isn't.

QUESTIONS FROM THE TRAIL

Ask Yourself

1. If your best AI user left tomorrow, how much of what they've figured out would stay?

2. When was the last time someone in your organization shared an AI experiment that didn't work — and what happened when they did?

3. Are you measuring what your organization is trying, or what it's learning?

TOOLS FROM THE TRAIL

A Shared Learning Log

One of the simplest tools I've seen work in this space isn't an AI tool at all — it's a shared log. A running document, Notion page, or dedicated Slack channel where team members capture what they tried, what happened, and what they'd do differently. Not a formal debrief. Not an ROI summary. Just a living trace of institutional knowledge, low-friction and ongoing.

This isn't an endorsement of any particular format, and it won't solve a culture where sharing feels risky. But it creates a visible structure for the behavior you're trying to normalize, and it makes learning searchable rather than perishable. It works best when a leader contributes to it actively — the format signals what's valued, and that signal has to come from somewhere.

ONE SMALL STEP

Navigating This Together

 If this pattern sounds familiar, I'd love to hear how it's showing up in your organization — just hit reply. 

And for a fuller look at the organizational patterns behind this, and the research on what learning cultures actually do differently, here's a link to this month's companion piece on Medium.

A quick note: You’re receiving this email because we’ve connected at some point about AI, leadership, or communications/marketing, and I thought you might find these field notes useful. I’ll send this about once a month. If it’s not helpful, you can unsubscribe anytime using the link below—no hard feelings.