When I was a kid, I watched my mom make beach plum jam in our Cape Cod home. She'd carefully skin each tiny fruit, separating the good part from the bitter part (most of the it!) She'd cook it down for what seemed like days, and taste it every hour or so. She knew something the recipe could't quite capture: how much water had to cook off to make it set, how much sweetness the sugar added and what was too much, when to finally pour it into jars to set. The knowledge lived in her, built through years of repetition.

She would have laughed at the idea of consulting an app to make jam. But I’ve wondered since what would have happened if she had relied on one long enough. Whether the knowledge in her hands would have stayed.

A Lancet study gave me an answer of sorts. A group of experienced physicians at four hospitals in Poland were quietly losing a skill – not through negligence, not through any failure of intention, but because AI had gotten very good at detecting polyps during colonoscopies, and the doctors started trusting it to do that seeing for them. When researchers looked at what happened to those same doctors’ performance without AI assistance, the adenoma detection rate had fallen by 20%.

They lost a skill they didn’t know they were losing. And I think this is a parable for something much broader; something happening right now inside the organizations I work with, often invisibly, often with the best of intentions.

AI is doing things that were unimaginable five years ago. It is also, quietly, doing something else: it is removing the daily repetitions that build certain human capacities. The capacities that are eroding aren’t the ones that show up on a resume. They’re the ones that make an organization function, that make people want to stay, that determine whether a crisis gets navigated or metastasizes.

I keep returning to five of them, because they’re the five my clients most need and most struggle to sustain.

Empathy

Empathy is the effort it takes to see and hear another person, and to let yourself be seen and heard. That is the heart of connective labor, and no machine does it on your behalf.

When Pew asked Americans and AI experts what worries them, 57% of the public said they were highly concerned about AI leading to less connection between people. Among the experts building it, that dropped to 37%. The people designing the tools worry least about the thing your employees and customers worry about most.

People stay where they feel seen and known. It is one of the Seven Human Needs From Work, and it shows up in your retention numbers long before it shows up in your engagement survey.

Empathy is not a mood. It’s a practice that atrophies when it isn’t used. Empathy is what we practice when we pause before saying or sending something and ask: does this actually meet the other person where they are? That pause, that small act of considering another human’s experience, is what’s getting skipped. Each time we skip it, something between people frays a little.

The courage to talk about hard things

No AI will have a hard conversation for you. It will help you draft one. It will smooth your language and soften your edges. But the moment of actually sitting across from someone and naming the thing that needs to be named – that requires a human being who has practiced doing it.

I see this deficit in leaders at every level. They don’t care less. The tolerance for discomfort has been trained down by years in organizations where difficult conversations were routed around rather than had. The culture of avoidance predates AI, but AI makes it easier to avoid, more plausible to hide behind polish.

The leader who can say, calmly and directly, “I need to talk to you about something that isn’t working,” and mean it, and stay in the room while the other person responds, is doing something that can’t be automated. It is also, in most organizations I’ve seen, something genuinely rare. The scarcity makes it valuable.

Accountability

The MIT cognitive debt study found something that surprised me: students who wrote their essays using AI reported the lowest sense of ownership over that work. They struggled to quote from it accurately. It wasn’t theirs, even when it technically was.

Ownership is what accountability is built on. You can only be accountable for something you’ve genuinely engaged with, made decisions about, stood behind. When authorship becomes diffuse (when it’s not quite clear who made this, who chose this, who decided this) accountability follows authorship into the fog.

The leaders I’ve seen build the most accountable cultures are the ones who do the visible work of owning their own choices. Not perfectly. Not without error. But visibly. They say “I made that call and it didn’t land, and here’s what I’d do differently.” That act of standing in full ownership of a decision is what teaches everyone around them that they can do the same.

Discernment

Microsoft’s own researchers found that the more confident a knowledge worker is in AI, the less critical thinking they apply. The mechanism makes sense: confidence in a tool reduces the instinct to question it. Questioning, genuine, rigorous questioning of what’s in front of you, is the foundation of discernment.

Discernment is not skepticism for its own sake. It’s the practiced ability to evaluate rather than accept. To notice what’s missing. To ask whose interests are served by a particular framing. To catch the thing that’s technically correct but contextually wrong.

This is a learnable capability that many organizations are quietly de-investing in every time they reward speed over rigor, output over quality, confidence over accuracy.

Emotional awareness

Last year, Pew Research asked two groups the same question: how worried are you about AI reducing human connection? Among the general public, 57% expressed concern. Among AI experts, it was 37%.

I think about the people I work with every day inside organizations, and I take seriously what they’re feeling. Something intangible is changing. The quality of presence in rooms. The sense that the person across from you is actually attending to you, not optimizing you.

Emotional awareness is the capability that keeps that presence alive. It’s knowing what’s happening inside yourself well enough to stay in contact with what’s happening in the room. It’s what allows a leader to notice when someone has gone quiet and to ask about it. It’s what allows a team to feel that it’s together, not just co-located.

It doesn’t develop by accident. It develops through attention, and attention has to be practiced.

The research on AI’s return on investment tells a sobering story – 95% of organizations reportedly getting zero return on enormous investments in these tools. I’m not surprised. AI is capable, but the organizations with the highest returns are going to be the ones that pair AI capability with human capability. That means leading with empathy, naming hard things, owning decisions, exercising judgment, and staying in emotional contact with the people around them.

These aren’t soft skills. They are, increasingly, our unique (and human) inherited advantage.

xo

Moe

P.S. Seats are still available for tomorrow's Masterclass for coaches, consultants, facilitators, and other service providers. Register for Solve Don't Sell: Get hired without getting salesy.​

Moe Carrick is the founder of Moementum, a leadership and culture consultancy. She has spent three decades helping companies build workplaces fit for human life.