AI is being sold to businesses as a way to improve margin. Many employees are embracing it for the same reason. The problem is that businesses usually mean profit margin, while employees mean margin in their lives.

Same word. Different dream.

That may become one of the biggest workplace tensions of the next few years.

I learned this long before AI.

Back when I worked in the high-pressure world of advertising as a copywriter and creative director, my art director and I would sometimes leave the building and go to Starbucks. Not to waste time. To create margin.

We knew that if we stayed in the office, with people constantly checking on us, asking for things, and wanting updates, we would not have the mental room to come up with the big ideas everyone wanted from us.

That coffee shop time was not a break from productivity. It was productivity.

Some of our most creative moments were not spent typing at a computer. They were spent leaning back in a chair, getting some distance, and talking and sketching our way toward a better idea. That’s something I think many organizations still miss, especially now in the rush to adopt AI.

When AI creates efficiency, who keeps it?

For businesses, margin means more output from the same people. Faster turnaround. Lower labor costs. Less slack in the system.

For employees, margin means less drudgery. Fewer late nights. More breathing room to think, recover, and have some life left at the end of the day.

Once AI creates efficiency, somebody decides where it goes. Back to the human being? Or right back into the machine of work?

Recently, I was able to get away. I had time to enjoy nature, spend time with family, and read something not related to work. I got caught up in the characters and story of a novel. When I came back, I felt refreshed and inspired.

And sadly, I also felt the need to justify that time by telling myself it gave me some really good ideas for work.

Somehow even breathing room can start to feel like something you have to justify.

Efficiency can quietly consume margin

If AI removes low-value tasks and gives people room for better judgment and deeper focus, that’s progress. But that margin can disappear in two ways. Management can fill the gap with more tasks, tighter deadlines, and leaner staffing. Employees can fill it themselves, because many of us have been conditioned to treat freed time as space for more work.

That may look like progress on paper, but in practice it can become just another way work expands to fill every available space. Parkinson’s Law applies to AI. When tools create margin, the instinct is to fill it.

Evidence is already mounting. A UC Berkeley study tracking AI adoption inside a real company found that even without management pressure, workers filled every hour AI freed up with more work. Deep-focus time fell and cognitive fatigue rose. There’s even a name for it: “AI brain fry.” Whatever you call it, it’s another example of a tool promising margin but quietly consuming it instead.

We’ve seen this before. Email was supposed to make communication easier. Smartphones were supposed to make work more flexible. They did both, but they also made work more constant and harder to leave behind. AI could easily follow the same path.

The real opportunity of AI isn’t just to use it to do work faster. It’s to decide what kind of margin is worth protecting.

Human margin is not waste

In my classes, we use Steven Johnson’s Where Good Ideas Come From. His study of innovators in history points to conditions that produce breakthrough ideas like liquid networks, adjacent possible, error, serendipity, and slow hunch. None of those happen easily when every minute is scheduled, measured, and filled. They need margin. Time for reflection.

I’ve seen this in my own life too. Some of my best ideas have come in places that don’t always look productive: the break room, between conference sessions, at a social hour, in the casual conversation between one thing and the next. That’s often where ideas connect.

Organizations say they want creativity, insight, and innovation. Then they build systems that leave no room for the very conditions that make those things possible.

You can’t squeeze people into breakthrough thinking.

Margin is not always waste. Sometimes it’s the condition that makes better work possible.

A better question for leaders

Some push back on this. They say pressure is the point. Constraint forces creativity. Urgency eliminates mediocrity. There’s evidence for it. Companies built on relentless intensity have produced breakthroughs that more relaxed organizations never did.

While AI is being sold as a tool to give time back, some of the very companies building AI post job descriptions glamorizing 70-plus hour weeks, or a 996 schedule. The technology that promises margin is arriving with a culture that demands you surrender it.

But that model tends to work in specific conditions: mission-driven people who opted in, often early in their careers, working on outsized problems they personally find worth the sacrifice. It also has real costs: attrition, burnout, and the quiet departure of experienced people who have other options.

Importantly, it misses what AI actually changes. A high-pressure model squeezes harder to get more. AI removes the need to squeeze people just to get routine work done.

The question isn’t whether to demand high performance. It is whether human qualities AI cannot replicate, such as judgment, creativity, and strategic thinking, flourish under constant pressure or require something different.

The companies that benefit most from AI over time may not be chasing maximum short-term output. I’d bet on the ones that use part of the gain to create better conditions for human performance: more focus, less drudgery, better decisions, more sustainable energy.

A healthier AI model could look more like defining work clearly, what done well looks like, and letting people keep some of the margin they create – for better work, and for more life.

Who gets the margin

The deeper issue isn’t that employers and employees want opposite things. Often, they both want better results and a sense that work is making a meaningful difference. Tension comes from a misunderstanding about how those outcomes are produced. Work culture often treats margin as waste to eliminate rather than the space needed to think, care, recover, and do meaningful work well.

This can lead to loss of motivation. Worker morale is meaningful. When people lose heart, productivity erodes. Eventually, the best people leave. AI didn’t create that misunderstanding.

The real negotiation happening around AI at work isn’t just about efficiency or adoption.

It’s about margin. Who captures it. Who benefits from it. Who gets the breathing room.

At the agency, I used to run during my lunch hour. It relieved stress, helped keep me healthy, and didn’t take away from family time. Anyone familiar with the creative process knows downtime matters. My subconscious mind kept working on client problems and projects. More often than not, I came back from those runs with new ideas for the work I was doing.

That doesn’t mean I never worked long hours. Big pitches and tight deadlines sometimes meant late nights, work after the kids were in bed, and Saturdays in the office. That came with the business. But there’s a difference between working hard when the work truly calls for it and treating constant overwork as proof of commitment.

After several years, my boss called me into his office. He said that my art director and I had the best work in the agency. Our work won creative awards, produced profit for our clients, and we always met deadlines while handling more clients and projects than the other teams.

Then he said, “But…” You run at lunch and go home at night.

He didn’t understand that the margin was part of what produced the results he was getting.

Shortly after that meeting, my art director and I both left for other opportunities.

The future of AI at work may not come down to the technology itself. It may come down to who gets the margin.

This post was drafted with the assistance of ChatGPT and Claude. The ideas, experiences, and opinions are my own.