AI Preparedness Starts with Understanding

Imagine, if you will, a world where every new employee arrives on their first day having already read every policy, procedure, operating manual, and training document in your company.

Every document has been memorized. Every step is followed precisely. Every instruction is carried out exactly as written. Sounds amazing, doesn't it?  Now imagine discovering that half the process isn't written down.

Welcome to the Tribal Knowledge Zone.

The place where work gets done every day, but nobody is entirely sure how.

I suspect most organizations end up there eventually, especially when they're growing. In the beginning, it doesn't feel like a problem. There are only a handful of people. Everyone sits close enough to ask questions. If someone isn't sure what to do, they lean over a cubicle wall, send a Teams message, or walk down the hall. The process lives in conversation more than documentation.

And honestly, that's not necessarily a bad thing. Humans are remarkably good at filling in gaps. We adapt. We improvise. We find work-arounds. We develop relationships. We learn which reports are reliable, which systems occasionally misbehave, and who to call when something goes sideways. Most employees aren't trying to break the process. They're doing exactly the opposite. They're trying to keep work moving despite the obstacles.

That's one reason I find some of the conversations around AI interesting. There seems to be an assumption that AI will arrive, instantly understand everything, execute flawlessly, and somehow replace the people doing the work today.

I don't see it that way.

In many respects, AI reminds me less of the experienced employee and more of the new employee. An incredibly capable new employee, to be sure. One who can read and retain enormous amounts of information in a fraction of the time it would take a human.  But it's still learning from whatever information it has been given.

That distinction matters.

Let's say you've spent twenty years driving between Tacoma and Seattle during rush hour.

You know where traffic usually backs up. You know which exits become parking lots after an accident. You know which alternate routes look good on paper but rarely save any time in reality.

Then there are the things you've learned simply through experience. Maybe there's an exit you take under certain conditions. Maybe there's a side street that works when traffic is moving slowly but not when it comes to a complete stop. Maybe you've learned that what the GPS recommends isn't always the fastest option because every other driver is receiving the exact same recommendation.

You didn't learn those things from a map. You learned them from doing the work. Now imagine trying to explain all of that knowledge to someone who has never driven the route before. That's where things get interesting.

Because most organizations have a Susan.

Susan knows which customer information is usually incomplete. She knows which report should be double-checked. If Department A says no, she knows to ask Department B for another solution. Susan knows which exception isn't really an exception because it happens nearly every week.

The documentation may say one thing. Reality may say another. Susan has learned the difference. The challenge isn't that Susan exists. The challenge is that the organization may not fully understand everything Susan knows.

That's what I mean by AI preparedness. Not selecting the software. Not writing the prompts. Not deciding which tool to buy. Understanding the work well enough that someone else, whether human or artificial intelligence, can successfully perform it.

How does the process begin? What information is needed? Where does that information come from? How do people know whether it's correct? What happens when information is missing? Which decisions require judgment? What exceptions occur so often that they aren't really exceptions anymore?

The answers to those questions are valuable whether you're preparing for AI, automation, Lean initiatives, an ERP implementation, or simply trying to grow the business.

In fact, that's the part I find most interesting.

We often start by asking which solution we need. Maybe the better question is whether we fully understand the problem in the first place. Because sometimes AI is the answer. Sometimes automation is the answer, or Lean. And sometimes the answer is surprisingly simple and costs very little. But all of those roads eventually pass through the same place.

Understanding.

The companies that benefit most from AI won't necessarily be the ones with the newest technology. They'll be the ones that have taken the time to make their work visible. After all, AI can only learn what we teach it.

The real question isn't whether AI is ready for your business.

The real question might be whether your business is ready for AI.

I think Step 1 is making Tribal Knowledge visible before it walks out the door.