Much of the conversation around Project Spero, with the proposed AI data center here in Spartanburg, has revolved around a few similar questions that we keep hearing as a framework for processing development in general. How many jobs will it bring? How much tax revenue will it generate? Will it strain our power grid? Will it draw too heavily from our water systems? What are the environmental impacts?
These are certainly necessary questions. They are practical, measurable, and tied to the immediate realities of governance and infrastructure. However, they are not the only questions worth asking, nor are they the origin of where our concern or attention should stem from, despite the competing marketing messages meant to shape public discourse.
Beneath the debates about megawatts and gallons per minute lies a quieter transformation that is harder to see but just as consequential. Projects like this do not simply add another industrial facility to the landscape. They introduce a new kind of presence into a place. They materialize intelligence.
For generations, land use in the Carolina Piedmont has followed recognizable patterns. Fields became suburbs, forests became highways, and rivers became reservoirs, while textile mills rose and fell. Logistics hubs replaced smokestacks. Each phase reorganized the landscape around a dominant economic logic… agriculture, manufacturing, distribution.
Now something different is emerging. Proposed AI infrastructure here in Spartanburg and throughout the Southeast of the United States does not primarily produce goods, textiles, or even physical services. Its purpose is to process cognition. To store, refine, and distribute decision-making capacity and contribute to the global chain of commodifying intelligence.
In effect, this all turns land into substrate for thinking.
This may sound abstract, but its implications are intensely material. Data centers are among the most physically demanding infrastructures ever built. They require enormous electricity flows, steady access to water for cooling, stable transmission corridors, and continuous connectivity. They generate heat that must be managed. They demand redundancy and resilience. In other words, they reorganize ecosystems to support continuous computation.
The Piedmont is not being asked simply to host an industry, but to sustain a new layer of perceived planetary intelligence to meet the resource needs of large language models. I think that changes the conversation.
When farmland became suburbia, we asked whether roads could handle the traffic. When distribution centers arrived, we asked whether zoning permitted increased truck traffic. But when intelligence becomes land use, the questions shift in both ontological and material ways that we’re not processing.
How much river becomes cooling capacity? How much forest becomes a transmission corridor? How much atmospheric stability becomes heat dissipation? How much regional resilience is redirected toward maintaining uninterrupted cognition?
Human systems do not float above ecological limits. They are embedded within them. AI infrastructure does not escape this reality at all; rather, it intensifies it.
What we are witnessing in places like Spartanburg is not simply economic development. It is the localization of a global cognitive metabolism. Decisions made in distant financial centers or algorithmic markets are beginning to rely on landscapes like ours for their material continuation.
The cloud, it turns out, is made of land.
This does not make projects like Spero inherently good or bad. But it does make them more consequential than the language of “jobs versus environment” suggests.
We are no longer deciding whether to permit another factory or mill. We are deciding whether this landscape will participate in sustaining planetary-scale computation, and it’s a different kind of civic choice.
It asks us not only to measure output and impact, but to reflect on orientation. What kinds of futures are we grounding here? What relationships between land, water, and intelligence are we normalizing? And perhaps most importantly… what forms of attention will this infrastructure train us to attend to (or be attended by)?
Because once intelligence becomes land use, the question is no longer only what we build on the land. It is what kind of world the land is being asked to think into being.