Construction AI Lab is a no cost sudyco® initiative focused on understanding how AI is actually showing up across the construction industry—what’s being used, what’s working, what’s not, and what’s coming next.

The Human Cost of AI

“The moment dehumanizing language enters a room, something toxic has entered with it.”

We are living through a moment in which people are being subtly, steadily, and increasingly described as problems to be solved by technology.

Maya Angelou once told a story about being at her home when she heard someone across the room making a racial and homophobic joke. She did not ignore it. She did not rationalize it. She did not wait to see whether anyone was offended. She walked over, told the person they needed to leave, and then explained why: it was poison.Her point was deeper than manners. She believed that words are not harmless. Words are things. The moment dehumanizing language enters a room, something toxic has entered with it. And if you allow it to remain, it spreads.

That story has been on my mind as I watch what is happening with artificial intelligence.Not because AI itself is poison. It is not. AI is a tool, and like many tools, it can be used in ways that help people or harm them. But the direction of much of the current AI movement increasingly carries a poison of its own: dehumanization.

Human judgment is being described as friction. Human variation is being treated as inefficiency. Human participation is being viewed as a bottleneck. Human labor is being framed as a cost to be eliminated. Human systems are being reorganized to fit the needs and logic of machines.When that happens, we are no longer just talking about automation. We are talking about dehumanization.

What Dehumanization Looks Like in the AI Era

Dehumanization does not always arrive wearing a cruel face. Often it sounds intelligent, efficient, and even visionary.It sounds like this:

People are too slow.
People make too many mistakes.
Human judgment is inconsistent.
The system would work better with less human involvement.
AI can manage the complexity better than people can.
Eventually people will only do the work that machines cannot yet do.

At first glance, these may seem like technical observations. But beneath them is a much more dangerous assumption: that the human being is becoming the least valuable part of the system.

Once that assumption takes hold, it changes everything. A person is no longer viewed first as a source of judgment, creativity, responsibility, relationship, and moral agency. Instead, the person becomes a unit of labor, a point of failure, a variable to be managed, a behavior to be predicted, or a remaining cost to be removed.

That is what dehumanization is. And just as Maya Angelou said, it is poison.

Why This Matters in Construction

Construction is not just a technical enterprise. It is a human system – a team of teams. Projects succeed or fail not simply because of plans, schedules, specifications, contracts, and software platforms. They succeed or fail because of the quality of the human system that uses those tools. Trust, communication, judgment, coordination, leadership, shared understanding, accountability, and problem-solving are not side issues in construction. They are the operating reality of every project.

If AI is used to support people, strengthen judgment, reduce administrative burden, improve visibility, and help teams coordinate more effectively, it is useful. But if AI is used in ways that reduce people to data inputs, push decisions beyond the capacity of teams to understand them, strip judgment from the field, centralize control, or measure success by how much human participation can be removed, then the technology is no longer serving the human system.

The human system is being reorganized to serve the technology. That is the line we must not cross.

The Poison is in the Assuptions

One of the great dangers of this moment is that dehumanization is often hidden behind positive language.

More Integration is always better.
More centralization is always progress.
More data always leads to better decisions.
The platform should become the system of record.
The platform should determine what deserves attention.
AI recommendations are inherently better than human judgment.
Greater visibility justifies reduced autonomy.
If the platform can coordinate the work, people should adapt to the platform.

But changing the label does not change the substance. If the underlying idea is that human beings are primarily obstacles to system performance, that is dehumanization. If the goal is a world where people have less judgment, less agency, less participation, less meaningful work, and less control over the systems shaping their lives, that is dehumanization. If progress is measured by how successfully human involvement can be minimized, that is dehumanization.

We need to say this clearly because many people feel that something is wrong, but they do not yet have language for it. They sense the direction. They hear the claims. They watch the acceleration. But because the argument is wrapped in technical language, they struggle to name what troubles them.

The right word matters. The word is dehumanization.

Not Every Use of AI is Dehumanizing — But the Trajectory Matters

A superintendent using AI to summarize meeting notes is not the problem. A project manager using AI to organize information is not the problem. An estimator using AI to support research is not the problem.The issue is not whether a person uses AI. The issue is what AI-enabled systems are gradually asking people to become. Do they help people think better, coordinate better, communicate better, and make better decisions? Or do they increasingly determine what deserves, attention, shape decisions before people make them, centralize authority, reduce autonomy, and gradually reorganize the human system around the needs of the technology.

The question is not whether AI becomes more capable. The question is whether people, teams, and organizations become more capable alongside it. That is the question. If we fail to ask it, we may gradually accept changes to our organizations that we never consciously chose.  

Not Every Use of AI is Dehumanizing — But the Trajectory Matters

What makes Maya Angelou’s story so powerful is that she did not wait.

She did not say, “Let’s see where this goes.”
She did not say, “Maybe they didn’t mean it.”
She did not say, “It’s only words.”

She understood that when poison enters the room, you stop it immediately. That is the lesson for this moment. When someone speaks of people as expendable in the march toward AI progress, we need to stop it. When someone treats human dignity as secondary to efficiency, we need to stop it. When organizations adopt AI without asking what it is doing to human judgment, accountability, trust, and agency, we need to stop and examine it.

When the future is described as one in which fewer and fewer humans are needed, we need to say clearly: that is not a neutral direction. That is a moral choice.

What Construction Leaders Should Do Now

This is not a call to reject all AI. It is a call to reject dehumanization. That means construction leaders should begin asking a different set of questions:

Does this use of AI strengthen the people doing the work or diminish them?
Does it preserve meaningful human judgment?
Does it increase capability or create dependency?
Does it improve relationships and coordination or weaken them?
Does it help the field, or does it mainly create more distant control?
Does it support the human system, or is the human system being reshaped to fit the tool?

These are not secondary questions. They are the main questions. Because once dehumanization becomes embedded in the language, strategy, and systems of an organization, it becomes much harder to remove.

We Didn’t Stop It At the Door

The poison is already present in subtle ways. Our responsibility is to recognize it before it become invisible.

The Construction AI Lab exists to help leaders see how AI is actually showing up — what is used, what works, what does not, and what comes next. But increasingly, I believe our work must also do something more explicit. We must help people recognize that a line is being crossed.AI should serve human systems. Human systems should not be sacrificed to serve AI. Human dignity, judgment, agency, and meaningful participation are not inefficiencies to be engineered out of existence. They are the very things that make organizations, projects, and societies worth having in the first place.

Maya Angelou understood that poison must be stopped at the door. That may be our task now as well.

To say, clearly and without apology:

AI should serve human systems.
Human systems should never be sacrified to serve AI

— Sue