Where leaders actually figure out AI.

I can usually tell within a minute of talking to a leader whether they feel good about AI. The ones who are calm have a tell. They reference a conversation. Something a peer told them, a thing someone in their network tried, a comparison they ran against people doing the same job. The ones who are anxious reference content. Articles they read, posts they saw, a vague and growing sense that everyone else has figured out something they have not.

That gap is not about intelligence and it is not about who reads more. It is about who has somewhere honest to put the question. The anxious ones are carrying it alone. The calm ones put it down in a room full of people like them, on a regular basis, and found out the truth was less frightening than the silence.

We lost the room where this used to happen.

For most of working history, you got educated about new things without noticing. You worked near people who were a step ahead, and the learning came through proximity. Over a coffee break. At the next desk. In watching how someone handled a hard moment and quietly filing it away. Nobody scheduled it. It was the ambient benefit of being in the same place as other people who knew things.

A lot of that is gone, and we have the data on exactly how it broke. When Microsoft studied 61,000 of its own employees through the shift to fully remote work, and published the result in Nature Human Behaviour, the finding was clean. People's collaboration networks became more static and more siloed, with fewer of the bridging ties that carry new information between groups. The casual cross-pollination did not move online. It mostly just stopped.

So the quiet way leaders used to absorb what was changing around them has thinned out at the exact moment the thing changing around them is the biggest shift in a generation. That is why mentorship is suddenly something everyone is chasing on purpose. It used to be free and ambient. Now you have to go and build it.

Why your brain reaches for peers under uncertainty.

Here is the part most people never learned, and it explains the whole thing. In 1954 the psychologist Leon Festinger described what we now call social comparison. His argument was simple and has held up for seventy years. When there is an objective way to measure yourself, you use it. When there is not, you reach for other people. And not just any people. You reach for the ones most like you, because their measure is the only one that actually maps to your situation.

Now look at AI through that lens. There is no objective yardstick yet. No agreed standard for how far along a leader in your role, in your industry, at your stage should be. The benchmarks do not exist. Which means the only honest measure available to you is the one Festinger named. Compare with people in your seat.

That is also why peer groups produce a very specific feeling on the way out. People walk in braced, certain they are behind. They compare notes with a room of genuine peers. And most of them leave lighter, because they discover they were measuring themselves against an imaginary standard, and against their real peers they are doing fine. That relief is not soft. It is social comparison doing exactly what it evolved to do, replacing a frightening guess with an accurate read.

Adoption moves peer to peer, not top down.

The other reason peers matter for AI specifically is that this is how technology has always spread. Everett Rogers spent a career studying how new ideas move through a population, and his diffusion of innovations work landed on something most leaders get backwards. People do not adopt a new technology because a study proved it works. They adopt it because someone like them, who already tried it, told them it was worth it.

Rogers was blunt about the mechanism. Most people do not evaluate an innovation on the science. They depend on the subjective evaluation of near peers who have already adopted it. Mass information creates awareness. It is the conversation with someone in your position that actually moves you to act, because their read carries a credibility no vendor deck and no headline can.

This is why reading more about AI does not fix the anxiety, and a peer conversation does. The article tells you the tool exists. The peer tells you whether it was worth the trouble in a business that looks like yours. One is information. The other is the thing you were actually missing.

The room only works if it is safe.

There is a catch, and it is the part people get wrong when they try to build this. A peer group only delivers if people stop performing in it. The whole value is the honest question. What are you actually using. Is it really helping or did you quietly shelve it. What did you measure going in, and did you hit it. Those questions only get asked where it is safe to admit you do not have it all figured out.

This is the most researched idea in team performance. When Google ran Project Aristotle to find what made its best teams work, the runaway answer was not talent, seniority, or who was in the room. It was psychological safety, a term from Amy Edmondson at Harvard, meaning a shared belief that the group is safe for taking an interpersonal risk. Safe to say I do not know. Safe to admit the thing I bought is gathering dust.

A peer group with that safety is worth more than any amount of reading. A peer group without it is a room of people protecting their image, which teaches no one anything. When you go looking for a group, that is the thing to test for, not the logos in the room.

Reading about AI tells you the tool exists. A peer who already tried it tells you whether it was worth it. Only one of those changes what you do on Monday. Chris Schafer

What a good peer group actually does.

Strip away the theory and the move is plain. Find a handful of people genuinely in your seat. A CFO in technology finds other CFOs in technology. A founder in manufacturing finds founders in manufacturing. Close enough that their experience maps onto yours. A local chamber, an industry group, a curated circle of peers, the form matters less than the fit.

Then use it for the concrete questions, not the abstract ones. Not "is AI going to change everything," which goes nowhere. The useful ones are specific and a little vulnerable.

• What are you actually using right now, not what you read about, what is open on your screen this week.
• Is it helping, honestly, or did it turn out to be a demo that did not survive contact with real work.
• What did you measure going in, and did you hit it, because a result with no metric behind it is just a feeling.

Run those in a safe room with real peers and you get the one thing no article can give you. An accurate, current read on where you stand, from people whose read actually counts. That is the stress test. It is cheap, it is fast, and almost everyone passes it better than they feared.

Where this gets built.

A peer group is the diagnosis, not the cure. It tells you where you stand and it calms the panic, which is no small thing. But turning that read into durable capability is a different job, and it is the one I do. We run it through HELP, the operating system my wife Elisha and I built. Hear, Evidence, Learn, Proceed. The peer group surfaces the gap. The mentorship closes it, with reps and feedback, the way you build any real skill.

Underneath all of it is the thing I have believed for thirty years. IQ only measures the ability to learn. The application of learning is EQ, and the human skills that decide who wins the AI transition, curiosity, communication, and confirmation, are EQ skills, and the through-line of the Conversation Intelligence work I write. A peer group is where you practice them in the open. That is the revenue leader mentorship, and it is why I keep telling leaders the same thing. Stop reading alone. Go find your room.

One question to sit with.

Ask yourself one thing. When the AI question gets loud in your head, where does it go? If the honest answer is nowhere, that it just sits there and compounds, that is the whole problem, and it has a simple fix.

The leaders who feel good about this are not smarter than you and they are not further ahead. They just stopped carrying it alone. Go do the same.

Last updated · June 2026