Your business runs on things nobody wrote down.

Imagine handing a new employee your company’s shared drive and telling them they have everything they need to do the job.

Technically, the information might be there. The price list. The procedures. The customer records. But by lunch, they would probably have questions. Which price list? Does that procedure still apply? Who can approve an exception?

Someone would lean over and explain how things actually work.

I think about that when I hear the pitch for connecting AI to a business. Give it access to your documents and tools, and let it get to work. There is real opportunity there. But access to the files is only part of understanding the job.

In the last issue, I asked what we could still accomplish with the AI we already have. One place I would start is with the things our businesses know but have never quite explained.

The customer who says “whenever you can” but has a launch next Thursday. The spreadsheet everyone uses even though a newer one exists. The discount that was a one-time exception and has quietly become an expectation. People carry that context between conversations. Much of it never reaches the official process.

AI may reveal how much of our work depends on things we assume someone else already knows.

My working theory is that making this knowledge usable will become a much more valuable part of work as AI handles more of the execution. We will need people who can surface the missing context, decide what matters, and explain it well enough for a colleague or an AI assistant to act on it.

Some of that can be written down. Some needs a conversation. And some requires a decision that nobody has made yet. Putting an old habit into a document doesn’t automatically make it a good rule.

Five skills, in practice

That puts familiar human skills at the center of a changing job. Companies are trying to anticipate what comes next: the World Economic Forum’s employer research highlights thinking, creativity and adaptability alongside technical skills. No forecast gives us a permanent checklist, but my bet is that emotional intelligence, critical thinking and judgment, communication, creativity, and adaptability will remain valuable as tools and roles change.

To make that practical, let’s look at a Learning team, a function I know well. Its work spans business relationships, project management, design, delivery and evaluation. These examples focus on how Learning professionals use the skills themselves. As you read, consider where the same needs show up in your business.

Emotional intelligence

AI changes more than a workflow. It can change how people understand their expertise, their contribution, and their future. Learning professionals need to recognize how that affects confidence, motivation, and willingness to learn, including their own. A business partner’s resistance might reflect concern about losing expertise or control. A colleague’s silence might conceal uncertainty. Recognizing those possibilities helps a Learning team ask better questions, build trust, and uncover concerns that a project plan or needs assessment might miss.

Critical thinking and judgment

When producing a polished learning solution becomes easier, deciding whether it should exist becomes more consequential. What business problem are we solving? Is it a skill gap, or an unclear process? What must employees understand themselves, and where can they use AI support? Learning professionals need to challenge assumptions, assess the evidence, and make those choices. Our judgment shapes what gets taught, what gets practiced, and whether we measure success by completed courses or improved performance.

Communication

As work moves between people and AI, the understanding that once traveled through informal conversations needs to become more explicit. Learning professionals must draw out what the business means, explain their recommendations, and make priorities, constraints, and expectations clear enough for others to act on. A request to “get people ready” needs a shared understanding of what ready means, who owns which decisions, and what the team can realistically deliver. Generating more messages is easy. Establishing that understanding and keeping it current takes deliberate communication.

Creativity

AI gives Learning teams more possibilities for what they can build and how people can practice. Creativity matters in recognizing what those possibilities could change. Could learners rehearse difficult decisions with feedback instead of reading another explanation? Could support be available at the moment someone needs to perform? The opportunity is to reconsider the learning experience itself. That requires imagination grounded in how people learn, what the work demands, and what would make the experience useful.

Adaptability

Adaptability needs to be built into the work. For Learning, that could mean a modular, skill-based curriculum that can change without being rebuilt; source content organized and labeled for AI ingestion; and learning structured around practice and application. Content and delivery systems should make it practical to update, reuse, and deploy learning as business needs evolve.

It also requires an adaptable mindset. Our roles, businesses, and tools may change, sometimes faster than we feel ready for. Learning professionals need to be willing to develop new capabilities, reconsider familiar methods, and revise plans they have invested in. Preparing for change means expecting to keep learning and making choices today that leave room to work differently tomorrow.

These skills work together. They help a team uncover what has gone unsaid, decide what matters, and adjust how the work gets done. Learning is one example. Every business has work that depends on the same capabilities.

The business has a job to do, too

These skills take more than individual effort to develop. Businesses need to give people time to experiment, practice decisions, and learn from feedback. Make decision authority clear, create room to question established approaches, and reward sound judgment alongside speed.

Start with a real piece of work. Have an experienced colleague explain what they noticed, why they chose an approach, and what would make them change it. Capture what others can reuse. That gives both the team and its AI better context.

I’ll explore that organizational support further in a future issue. For now, ask: What would a capable new colleague get wrong here because nobody thought to tell them?