Every business has that person. They know a little about marketing, enough about technology to ask uncomfortable questions, and more about your customers than their job description suggests. They can look at a website, a sales process, or an internal workflow and tell you something is off. They might not always know how to fix it themselves.
AI will widen the gap between employees who take ownership and employees who wait to be told what to do.
That is where AI gets interesting. I think some of the people best positioned to use AI are the jacks of all trades who have spent years developing something harder to teach than a software shortcut: taste.
Not taste as in choosing a nice font. Taste as in knowing what good looks like, understanding why something is wrong, and recognizing when a technically correct answer still misses the point.
These people have often been limited by execution. They can see the better version, but building it requires skills they have not mastered, time they do not have, or three other departments to get involved.
AI can help close that gap. Someone who understands your business can use it to explore an idea, draft the copy, sketch the workflow, analyze a spreadsheet, or build an early prototype. Work that previously stopped at “we should do something about this” can start becoming something people can review.
That does not make the first version good. It makes the first version possible. The valuable part is what happens next.
Your generalist looks at the output and says, “That sounds nothing like us.” Or, “Our customers would never use that.” Or, “You solved the marketing problem and created an operations problem.”
Then they adjust, test, and try again. That is not just prompting. That is judgment applied repeatedly.
And judgment matters because AI can make weak ideas look surprisingly finished. Clean formatting, confident language, and a working demo can hide bad assumptions. Someone still needs to notice that the beautiful dashboard measures the wrong thing.
The person who has worked across your business has an advantage here. They understand that the website is connected to sales, sales is connected to delivery, and delivery is connected to whether customers ever come back. They see the handoffs.
A specialist might improve one part of the business. A capable generalist can recognize when that improvement creates friction somewhere else. Give that person more execution capacity, and their breadth becomes a practical advantage instead of an awkward answer to “What exactly do you do here?”
But let’s not turn this into another excuse to pretend expertise no longer matters. Knowing enough to spot a problem is not the same as knowing enough to approve the solution. Your generalist still needs specialists, particularly when the work involves security, sensitive data, financial decisions, or systems people depend on. Good taste includes knowing when to call someone who knows more.
For companies, the opportunity is not simply handing everyone an AI subscription and hoping productivity happens. Find the people who already connect the dots. Give them a real problem, appropriate access, clear boundaries, and room to experiment.
Ask them to show their work. What changed? What did they check? What still needs a human decision? Did the result help? Start with something small enough to test and useful enough to matter.
You may discover that the employee you thought was “pretty good at a lot of things” was never short on ideas or judgment. They were short on the capacity to execute all of them.
AI does not give them taste. They earned that. What it can give them is more ways to put that taste to work.
That is a person worth investing in.

