By Sahver Kaya · 3 min readYou build AI skills by giving people a private place to ask, try and get it wrong, a few minutes a day, with nobody grading them. Most people already have an AI question they have not asked, because asking out loud says "I am behind."
RCM ThinkLabs (rcmlabs.io) gives people that place: short sessions of private practice where a mistake costs nothing. It is run by a game-theory engine built on research at MIT.
Why do employees stay quiet about AI?
Because asking looks like falling behind. Everyone assumes everyone else has it figured out, so the real gaps stay hidden, and they are usually the useful ones.
The person who doubts an answer is often right to. That doubt is the start of good AI use, and it rarely gets said.
Why does a workshop not fix it?
A workshop puts people in a room with their manager and their peers, so the embarrassment walks in with them. They nod, take notes, and go back to their desks with the same question.
It is one reason AI adoption stalls after the rollout. The tool is there, and the question is still unasked.
What is a safe place to practice AI skills?
A safe place is a private space where nobody is graded, a wrong move costs nothing, and you can try the same thing again tomorrow. It is psychological safety made practical, and short enough, a few minutes a day, to fit inside the job.
Each person comes to it alone, with their own lens, because an analyst and an engineer read the same situation differently. Then the team compares notes, and that is where people see what they could not have seen alone.
Which skills decide how well people use AI?
Power skills do. Four come up first.
- Clarity. Saying what you want before you prompt.
- Critical thinking. Checking what comes back, including the answer that sounds right.
- Systems thinking. Seeing where the answer fits and what it changes downstream.
- Active listening. Hearing what the other person needs, so the prompt starts from the right question.
The list does not stop there. Leadership, persuasion and every other skill the AI era asks for belong on it too.
How do you know it is working?
You see people ask more questions. At an advanced engineering team working with the U.S. Air Force, the share of questions that asked why went from 1 in 17 to 1 in 7 over 39 days.
That is one habit, read against each person's own starting baseline. We did not measure AI tool usage, and we make no claim about it. But a team that asks why is a team that checks, and a team that checks uses AI better.
What can a manager do this week?
Say out loud that you do not have it figured out either. It gives everyone else permission.
Common questions
How do you build AI skills when employees are afraid to ask questions? Give them a private place to practice, a few minutes a day, where nobody is graded and a mistake costs nothing. Then let the team compare notes.
How do you create psychological safety around AI adoption? Make asking cheap. Practice stays private until the person chooses to share, and managers say out loud that they do not have it figured out either.
Is a safe place just a chat channel? No. Channels reward the people who already feel confident. A safe place means nobody is graded and nothing is shared until the person chooses to.
How long before it shows? Habits move in weeks. We saw the share of questions that ask why rise from 1 in 17 to 1 in 7 inside 39 days.
Does this replace instruction on AI tools? No. Tool instruction teaches what the buttons do. This builds the skills that decide how well the buttons get used.
Is free time to explore enough? Time without feedback rarely turns into skill. Short, repeated practice with feedback does.
See it on your own team.
Sahver Kaya is the founder and CEO of RCM ThinkLabs. An educator, experienced builder, and MIT alum, she is driven by one conviction: artificial intelligence, used well, should make people sharper.
Connect on LinkedIn →