Fast and Lonely

In short: AI lets people get answers without asking colleagues. That saves time but removes the small conversations where trust forms, so teams can become faster and more isolated at the same time. Leaders should protect some human contact on purpose.

A few years ago, if I needed to understand something about a client's platform, I'd walk over to a developer and ask. The answer took ten minutes. The conversation took twenty, because it drifted: how the project was going, what was frustrating them, a story about the last time this client changed their mind. I walked away with my answer and something else I didn't know I needed.

Now I can ask a model. The answer takes thirty seconds. It's often better organized than what the developer would have told me. And I never have the conversation.

Multiply that across a team, across a year, and something changes that doesn't show up in any productivity metric.

Why Do Inefficient Conversations Matter?

Much of what makes teams work happens in moments that look like waste. The question that turns into a conversation. The review that turns into mentoring. The status meeting that runs long because someone mentions they're struggling. The quick favor that builds a debt of goodwill someone will repay months later.

None of that is efficient. All of it is how trust forms. People come to rely on each other by depending on each other, in small ways, repeatedly. Each small dependence is a chance to be helped, to help, to discover who is generous and who is careful and who knows things nobody wrote down.

AI removes many of those small dependences. Why ask a colleague when the tool is faster, available at midnight, never busy and never annoyed? Each individual choice makes sense. Together they thin out the web of contact that holds a team together.

How Does Asking for Help Build Trust?

Asking someone for help is a small act of vulnerability. You admit you don't know something. They choose to spend time on you. That exchange, repeated, is the raw material of trust.

A team where nobody asks anybody anything can still function. Work gets done. Output may even rise. But the relationships go untested. When the hard moment comes, a missed deadline, an angry client, a disagreement about direction, the team discovers it doesn't really know each other. There's no reservoir of goodwill to draw on, because nobody built one.

This isn't only a workplace problem. In 2023 the U.S. Surgeon General issued an advisory calling loneliness a public health epidemic, and that trend began long before generative AI. But AI makes it easier to go a whole workday without needing anyone, and needing people turns out to be part of how we stay connected to them.

What Are the Signs a Team Is Getting Lonely?

Questions disappear from shared channels. People stop asking in public because they can ask a tool in private.

Juniors stop seeking out seniors. The informal mentoring that happened through questions dries up, and nobody notices until growth stalls.

Meetings get shorter and colder. Everyone arrives having already done their analysis alone. There's less thinking together, more presenting at each other.

Conflict surfaces late. Without small, regular contact, disagreements don't get worked out early. They build up and arrive all at once.

Isn't Asking AI Sometimes Better?

Not everyone misses the old way. For some people, especially newer or more introverted team members, asking a tool is a relief. They can ask the basic question without feeling judged, at their own pace, without interrupting someone busy. That's a real benefit, and it shouldn't be dismissed. Remote and hybrid work had already thinned out casual contact long before AI arrived, too.

The point isn't that every question should go to a person. It's that some should, on purpose, because the asking itself builds something. If AI handles the questions people were afraid to ask, the time it frees up can go toward the conversations that matter more. That only happens if someone decides it should.

How Can Leaders Protect Human Contact?

The answer isn't to make people ask colleagues questions a tool could answer better. It's to protect the human contact deliberately, now that it no longer happens by accident.

Think together before using the tool. For important work, start with a conversation among people. Use AI after, to extend and test what the group came up with, not to replace the discussion.

Keep some work collaborative on purpose. Pair on problems. Review work in person. Not because it's faster, but because the relationship is part of what you're building.

Make asking normal. Leaders who ask questions openly, including ones they could have looked up, signal that needing help is fine. That matters more now that nobody strictly has to.

Protect the drift. The conversation that wanders off topic is where people learn who their colleagues are. Leave room for it in meetings and in the week.

Notice who goes quiet. In a team that depends less on each other, someone can disappear into their tools for weeks before anyone notices they're struggling. Somebody has to be paying attention.

What We Trade

Every efficiency is a trade. Usually the trade is worth it. But some of what AI saves us from was never really waste. It was the slow, inefficient, irreplaceable work of becoming people who can count on each other.

A team can be very fast and very lonely at the same time. The speed shows up in the numbers. The loneliness shows up later, in the moment the team needs each other and finds out it never learned how.

Frequently Asked Questions

Does AI make teams lonelier? It can. When people ask a tool instead of a colleague, they lose the small exchanges that build trust. The effect is gradual and rarely shows up in productivity metrics.

Is it bad to ask AI instead of a coworker? Not always. For basic questions, AI can be a relief, especially for newer or introverted team members. The goal is to keep some questions going to people on purpose.

How can teams stay connected while using AI? Think together before using the tool on important work, keep some work collaborative, make asking for help normal and notice when someone goes quiet.

Evan Prout is a strategist in Grand Rapids, Michigan, who writes about how organizations think and decide with AI.