By Dominick Miserandino, CEO and Founder, RTM Nexus
There’s an interesting problem developing in corporate America. CEOs are spending millions on AI, hiring consultants to explain it, and telling employees they need to adapt. Yet I wonder how many of those same executives have actually sat down and used the technology beyond asking ChatGPT to summarize a report.
I’m not talking about attending an AI conference or watching a demonstration. I’m talking about getting into the software, making mistakes, figuring out what works, and understanding why something that looks simple can become surprisingly complicated.
I’ve been a CEO, a CMO, a developer, and just about everything in between. One thing I’ve learned is that you can’t effectively manage people if you don’t understand the mechanics of what they’re doing.
When I was CMO of AdoramaPix, which printed photographs, I went into the production facility and learned how to print photos. I worked with the presses. I even delivered orders.
I wasn’t planning to become a printing technician. I wanted to understand the operation well enough to make better decisions about it.
AI requires that same approach.
Understanding the work changes the decisions
Consider a CEO who announces that AI should reduce the amount of time employees spend on administrative work by 30%.
Sounds reasonable. Who doesn’t want greater efficiency?
But what happens when the technology retrieves the wrong information, duplicates records, or makes a decision based on outdated data? What happens when the employee spends more time correcting the work than doing it themselves?
Those are things you discover through actual use.
I’ve spent considerable time building AI agents that handle different functions. Some check databases. Others research information or update records. I’ve learned that a seemingly simple instruction can create all sorts of unexpected complications.
The important part is understanding where the complications originate.
Is the information wrong? Are the instructions unclear? Is the system accessing the correct database? Does it know when to stop and ask a question?
These are operational questions. They’re also management questions.
An executive who has never worked through them may see an employee struggling and conclude that the employee is resistant to change.
The employee may simply understand the technology’s limitations better than the person directing them to use it.
Every department speaks a different language
Having worked in development, marketing, sales, and executive management, I can tell you that each department has its own language.
A developer looks at how a system functions. A marketer considers how it affects customers. A salesperson thinks about whether it helps close business. A CEO has to understand the financial implications of all three.
AI doesn’t eliminate those differences.
It can actually make communication more difficult when management assumes the technology solves a problem without understanding how that department operates.
You don’t need to become an expert programmer to lead an AI initiative. But you should know enough to recognize when a proposed solution is practical, when it’s adding unnecessary complexity, and when the numbers being presented don’t reflect what’s actually happening.
Start with something real
My advice to executives is fairly simple.
Take one repetitive task you personally understand and try to automate part of it.
Maybe it’s reviewing weekly sales reports. Maybe it’s comparing customer information across multiple spreadsheets. Maybe it’s finding inconsistencies in inventory records.
Start small. Look at the results. Find the mistakes. Try to improve it.
Then ask your employees what they’re experiencing.
You’re going to have a much more productive conversation when you understand the problems firsthand.
And don’t assume that because something works once, it’s ready to become company policy. Running an experiment on ten records is very different from applying the same process to 100,000 customer accounts.
The financial consequences change considerably.
Businesses have survived countless technology transitions. I’ve worked through the internet, search, social media, mobile technology, and e-commerce. In every case, the executives who understood the underlying business operations had an advantage over those simply following the latest trend.
AI is moving quickly, and nobody knows exactly where it will end up.
That’s fine. Leadership has never required knowing everything about the future.
It does require understanding what you’re asking other people to do.
And before committing the company to a major AI transformation, perhaps the CEO should spend a few afternoons actually using it.
I discussed this approach to AI leadership on the Proteus Leader Show, including why firsthand experience matters when managing technological change.