As AI transforms the infrastructure behind our digital lives, architecture may need to rethink not only how data centers are built, but how they belong to the places around them.
Architecture changes when the function changes.
For much of the 20th century, the office was organized around the workstation: a desk, a defined place for each person, a clear separation between work and everything else.
Then the nature of work began to change.
Technology companies took that transformation much further. Their campuses were no longer simply places where people went to work. They became environments for collaboration, concentration, interaction, food, culture and community.
And something else became clear.
A campus could not simply function as an island that filled up in the morning and emptied at the end of the day. Its relationship with the surrounding place became part of the way it was conceived.
The place itself became part of the information needed to design it.
The workplace changed because the work changed.
Today, the difference is not only the scale of the transformation, but the speed at which the environment around it has to adapt.
We used to design places to house people who thought.
Now we are designing places to house machines that think.
But those machines do not exist outside the places that support them.
They depend on land, energy, water and infrastructure, all embedded within places where people live, work and move.
Unlike a campus, very few people work inside a data center. Yet its presence can affect the daily life of many people who have no relationship with the work happening inside it.
Its demands can extend beyond its physical boundaries.
Energy infrastructure. Water resources. Transmission networks. Roads. Land use. Noise. Employment. Economic development. The relationship between infrastructure and community.
The building may be highly specialized.
Its consequences are not.
And this raises a different architectural question.
Maybe the question is not only how to build a better data center.
It is whether we should rethink the relationship between this new infrastructure and the place around it in the same way we once had to rethink the workplace.
If the function has changed this radically, should the conception of the place change with it?
The history of architecture is, in many ways, a history of this relationship.
When the way we live changes, the spaces that support us change.
When the way we work changes, the workplace changes.
And when the infrastructure supporting intelligence changes, perhaps the places that contain it must change too.
The data center is often treated as a highly technical object: a building optimized for computation, cooling, power and security.
But AI is changing the scale and significance of what happens inside those buildings.
These are no longer simply facilities supporting the digital economy.
They are becoming part of the physical infrastructure of an increasingly intelligent one.
That distinction matters.
Because once infrastructure becomes this consequential, architecture cannot stop at the building envelope.
It has to consider the relationship between the building, the systems that sustain it, and the place in which it exists.
Perhaps the next paradigm is not simply a better data center.
Perhaps it is a different conception of what that infrastructure can become.
Not an isolated machine.
Not simply a technical facility.
But infrastructure that understands its relationship with the place around it.
Because when the function changes radically enough, perhaps the architecture should change with it.
