AI Will Not Fix a Poorly Designed Workplace

I like AI.

I use it every day.

It can help people synthesize information, generate ideas, reduce repetitive work, and move more quickly from a blank page to something useful. It is one of the most powerful workplace technologies introduced in our lifetime.

But I think we are missing something important in the conversation about AI at work.

We are introducing this technology at a moment when many people are already burned out, lonely, disconnected, and struggling with the cost and pace of modern life. Organizations are understaffed, under-resourced, and under enormous pressure to produce more.

Then we introduce AI and say, “Great. Now we can do even more.”

But is more output the only opportunity?

Productivity is a choice, not an automatic benefit

When a tool saves time, leaders make a design decision about what happens to that time.

If AI gives an employee five hours back each week, those hours can immediately be filled with additional tasks. The performance baseline can rise. What was once considered exceptional can become the new minimum. The organization may produce more without reducing pressure or improving anyone’s experience of work.

That is one possible design.

But reclaimed capacity could also create more time to think. More time to mentor. More time to collaborate, learn, build relationships, check quality, and solve difficult problems. It could allow a chronically overloaded team to move from unsustainable to manageable. It could make room for the deeply human work that constant busyness keeps pushing aside.

AI does not decide between these possibilities. Leaders do.

The technology creates capacity. Workplace design determines who benefits from it and how.

Technology can reinforce what already exists

Organizations often treat a new tool as though it will automatically improve the workplace. But technology tends to amplify the priorities, habits, and power structures surrounding it.

In a thoughtful culture, AI may support learning, access, creativity, and better service. In an urgency-driven culture, it may help people move even faster without questioning whether the pace is wise. In an understaffed organization, it may help a team keep producing without addressing why adequate staffing remains out of reach. In a disconnected workplace, it may automate the few interactions that still brought people together.

This is the danger: technology can make a poor design more tolerable and therefore more durable.

If AI helps employees survive impossible workloads, leaders may lose the pressure to correct those workloads. If it produces polished communication quickly, teams may overlook the decline of direct conversation. If it becomes the first place people turn for ideas, organizations may unintentionally weaken reflection, collaboration, and confidence in human judgment.

Efficiency is useful. But efficiency is not the same thing as health, quality, meaning, or effectiveness.

Ask what should remain human

The best question is not simply, “What can AI do?” It is also, “What should people continue to do, and why?”

Some work has value beyond the final output. Mentoring develops people. Collaborative problem-solving builds trust. Writing clarifies thinking. If we automate only for speed, we may remove experiences that help people learn, connect, and make meaning.

Many administrative tasks deserve to be simplified or eliminated. The challenge is distinguishing friction that adds no value from effort that develops human capacity.

Leaders should ask:

• Which tasks are repetitive, draining, and appropriate for AI support?
• Where is human judgment essential?
• Which activities build relationships, knowledge, or confidence through the process itself?
• How will we protect privacy, quality, and accountability?
• If time is saved, how will that capacity be used?
• Who gains from the productivity increase, and who carries the new risks?

These are workplace design questions, not merely technical ones.

AI adoption is also culture change

Employees are not encountering AI from the same starting point. Some are excited and experimenting daily. Others are anxious about job security, skeptical about accuracy, concerned about ethics, or unsure what use is permitted. Some may hide their use because they fear judgment. Others may rely on it without understanding its limits.

If leaders focus only on licenses and training modules, they will miss the emotional and relational dimensions of adoption.

People need clear expectations. They need meaningful opportunities to learn. They need permission to ask basic questions and raise concerns. They need transparency about how AI may affect roles, performance standards, evaluation, and staffing.

Most importantly, they need evidence that the organization is using AI to create better work, not simply to extract more work.

Decide what you believe work should be

AI is often described as a force that will determine the future of work. But that framing can make organizations feel passive, as though the technology itself will decide what happens next.

Will increased capacity create breathing room or a higher treadmill speed? Will automation free people for connection and creativity or remove more human contact? Will AI deepen access to knowledge or concentrate decision-making? Will organizations use it to address poor processes or avoid harder conversations about staffing, priorities, and culture?

AI is not just changing how we work. It is exposing what we value about work and giving us an opportunity to redesign it.

So perhaps the central question should not be, “How can AI make us more productive?”

It should be:

What kind of workplace are we using AI to create?