Stop Hiring AI Natives. Train the People Who Already Know the Work.
Your workforce is your biggest competitive advantage. Not AI.
I sit in a weird spot.
Some weeks I'm in a conference room with executives - sharp, brilliant humans with 20+ years of industry experience and a vision. They see where AI could reshape operations, make processes leaner, let their teams do more with less. Good ideas. Big ideas.
Other weeks I'm sitting next to someone on the front lines who's been doing the work manually for 20 years. She's looking at the same tool the executives are excited about and trying to figure out where it fits into the actual rhythm of her day - the workflow she's built, the edge cases she's memorized, the client quirks nobody wrote down.
The gap between those two rooms is enormous. And I promise it is not being closed by firing the 20-year veteran and replacing her with a cheaper "AI-native" hire.
The only thing staff wanted during a recent AI rollout
In a recent AI enterprise roll out we got 1 piece of feedback from staff.
They just wanted a demonstration of the new platform.
They wanted trainings.
They didn't want to hear about the vision or the mission of the integration, they just wanted to get their hands dirty and start practicing.
Gallup just did a study with 15,066 U.S. workers, conducted with Amazon Gallup -they found that 57% of workers are very or extremely interested in training to upgrade their skills, and when that training is free and paid during work hours, interest jumps to 71% Amazon. 48% said they would switch jobs for skills training opportunities
The payoff for the ones who do upskill: an 8.6% average salary bump - about $8,000 more per year. 71% report greater job satisfaction, 30% moved into new, higher-paying jobs, and 39% advanced in their current role.
SO TELL ME WHY:
Only 2% of CHROs strongly agree their upskilling efforts are developing the skills employees need for the future Gallup.
In 2024 ( I can only assume this would jump to 99% in 2026 considering there are hospitals trying to implement AI to replace Cardiothorastic Surgeons PMC) 47% of employees strongly agree they have the skills to be exceptional at their current job - and only 2% of CHROs strongly agree their upskilling efforts are developing the skills employees need for the future. Two percent. Meanwhile, 22% of workers worry their job will be made obsolete by technology (again I'm sure this is 99% in 2026), and 72% of Fortune 500 CHROs expect AI to replace jobs in their organization within three years Gallup.
So workers are watching their leaders say out loud that AI is coming for their jobs, while those same leaders aren't actually teaching them how to use it.
Gallup's 2025 CHRO research shows the problem widening, not closing: 59% of CHROs now say employee development is the area they find most challenging while only 45% of U.S. employees took part in any training for their current role.
The business case isn't soft. It's P&L.
Gallup's own analysis of their engagement data: doubling the number of employees who feel they have strong opportunities to learn and grow could produce an 18% increase in profitability and a 14% improvement in productivity
And replacing the people you didn't train is expensive.
SHRM puts employee replacement at six to nine months of salary; Gallup puts it at 50% to 200% of annual salary.
For senior specialists, HBR estimates losing them can cost up to 20x average recruitment and training costs because of the institutional knowledge walking out the door - what HBR calls the "deep smarts" that only accumulate through years of doing the work.
The top-down mandate is the wrong playbook.
Here's where I'd push further than most of the research: the best ideas about how AI should actually be deployed inside your company will not come from an executive strategy offsite.
The best ideas about how AI should be deployed in your company will come from the person who's been doing the work for 20 years.
She knows which 40 minutes of her day are genuinely wasted. She knows which "edge case" happens 200 times a year. She knows the client who always emails Friday at 4pm because she figured out the pattern back in year three.
If you give that staff person training, give her space to experiment, give her paid time to collaborate with other colleagues accross the company or accross the industry - she will be the one to show you where the gaps are. The processes that can be automated. The systems that could be driven by AI so she can spend her time on the things that only she can do.
An executive with a Miro board of workflow diagrams doesn't have that. An AI-native hire fresh out of school doesn't have it either. I promise.
Top-down AI mandates fail because they're abstract. Bottom-up AI adoption works because the people driving it are solving problems they actually feel.
Gallup's research backs this up from another angle: 39% of employees report that role-specific skills training would be most beneficial to their development
Who do you invest in?
When training budget shows up, it tends to flow to the loudest person in the meeting, the most politically connected person, or whoever volunteers first.
Don't do that.
The people to funnel AI training, conferences, and unstructured exploration time to are the quiet high performers. The ones who consistently get the work done. The ones whose names come up when something hard needs to get done. The people who - when you ask "how do you handle X?" - have an answer that reveals they've been quietly improving their own workflow for years without being asked.
Give three of those people real AI training, genuine permission to experiment, and unstructured time to figure out how this fits into their actual work. Then listen to what they build.
One of them will come back in six weeks with something that saves your team hundreds of hours. That's worth more than ten executive AI strategy offsites.
It is not difficult and it's not expensive.
Start this week.
Reach out to a few high performers.
Have a candid conversation.
Ask them to take some time during the work day over the next few weeks to do some research.
Connect them with colleagues.
Ask them to experiment.
Share some ideas the C-suite is bouncing around.
Ask for their feedback.
Listen to their feedback.
Just like that, I saved you 10s of $1,000s of dollars in generic theoretical consulting fees.
Executives set direction and remove blockers. They don't design workflows.
Frontline experts get training, time, and permission to reshape their own work. They do design workflows.
Managers translate between the two - connecting the strategic "why" to the operational
"how."
(And here's a Gallup data point worth sitting with: fewer than half of managers worldwide have had any formal management education. That's a separate problem, but it's the same problem. Untrained managers can't coach AI adoption because they were never taught to coach anything.)
Your workforce is not a cost center to optimize away.
It is the one thing your competitors cannot copy.
The AI models are mostly the same. The prompts are public. The integrations are available to anyone with a credit card, including your compettition.
Your employees?
They are the ones that set you apart.
They are the ones with ideas that AI could never put together.
The employee who knows your customers by name and has been solving problems and building systems for decades.
Train her. Listen to her. Don't fire her.
Comments