TL;DR
Wired reports that an international buildout of costly data centers is accelerating as companies race for AI advantage. The piece frames those facilities as a growing physical and financial footprint — compared by some to an empire — and links the trend to rising costs and resource debates.
What happened
Lauren Goode’s Wired article describes a rapid expansion of large, expensive data centers tied to the competition to develop advanced AI. The story uses an analogy invoked by OpenAI’s CEO — comparing the company’s reach to a Roman Empire — to illustrate how AI operators are effectively acquiring global infrastructure. These facilities are cast as modern latifundia: concentrated parcels of capital and land used to host intensive computing. Wired frames the trend as both costly and wide-reaching, saying the AI arms race has left a substantial footprint that continues to grow. The article sits alongside related reporting that flags operational issues tied to the expansion — including debates over energy and water use and firms repurposing existing industrial-scale data halls (for example, bitcoin miners shifting toward AI workloads) — though specific project-by-project details are not enumerated in the piece.
Why it matters
- Consolidation of compute could concentrate economic and technical power among a few large operators.
- The rising cost of building and operating hyperscale AI facilities affects investment and industry structure.
- Expansion of data centers raises resource and infrastructure questions, including energy and water demand.
- Geographic deployment of these sites has geopolitical and local community implications.
Key facts
- Article title: 'Billion-Dollar Data Centers Are Taking Over the World'; author: Lauren Goode; publisher: Wired; published Dec. 28, 2025.
- Wired frames the proliferation of large data centers as a direct consequence of competition to develop AI.
- The piece quotes an analogy from OpenAI’s CEO likening the company’s reach to a Roman Empire, used to illustrate territorial expansion of infrastructure.
- The article describes these facilities metaphorically as modern latifundia—concentrated holdings of land and capital for AI compute.
- Wired characterizes the trend as leaving a growing physical footprint and becoming increasingly expensive.
- Related reporting cited alongside the story highlights discussions about data center water use and environmental concerns.
- Related coverage also notes some bitcoin-mining operators are repurposing industrial-scale facilities toward AI workloads.
- The article does not provide a list of specific new facilities, dollar amounts per project, or a comprehensive map of operators.
What to watch next
- Cost trajectories for building and operating hyperscale AI data centers, and how they affect industry entrants and incumbents.
- Local and national regulatory responses to data center siting and resource use (not confirmed in the source).
- Which companies expand into new regions and how that reshapes market concentration (not confirmed in the source).
- Outcomes of debates over water and energy impacts of large-scale AI facilities, including any policy or community actions.
Quick glossary
- Data center: A facility that houses computer systems and related equipment for storing, processing, and distributing data.
- Hyperscale: A design approach and scale of computing infrastructure intended to efficiently support very large workloads, often used by major cloud and AI providers.
- Latifundia (metaphor): Historically, large landed estates; used metaphorically in the article to describe concentrated holdings of land and capital for data centers.
- AI workload: Computational tasks required to train or run artificial intelligence models, often demanding high-performance processors and large amounts of power and cooling.
- Bitcoin miner (context): Specialized operations that validate cryptocurrency transactions using intensive computing; some operators have been reported to repurpose facilities for AI tasks.
Reader FAQ
What is the central claim of the Wired piece?
The article argues that an international expansion of costly, large-scale data centers is under way as companies race to build AI advantage.
Are these facilities really 'billion-dollar' projects?
The article’s headline describes them as billion-dollar data centers; the text frames them as large and expensive but does not list exact project costs.
Is OpenAI literally building an 'empire'?
Wired cites OpenAI’s CEO using a Roman Empire analogy to describe reach; whether OpenAI is 'building an empire' beyond that rhetorical comparison is not confirmed in the source.
Are environmental harms from data centers proven?
The story and related reporting note debates about energy and water use, but specific causal findings or quantified harms are not provided in the source.
Are bitcoin miners switching to AI computing?
Related coverage mentioned in the article indicates some large bitcoin-mining operators are repurposing facilities toward AI workloads.
LAUREN GOODE BUSINESS DEC 28, 2025 6:00 AM Billion-Dollar Data Centers Are Taking Over the World The battle for AI dominance has left a large footprint—and it’s only getting bigger…
Sources
Related posts
- Travel agents took a decade to collapse — software developers three years in
- Say No to Palantir in the NHS — Opposition Campaign and Call to Action
- Yanis Varoufakis on the Future of Capitalism, Cloud Capital and Tech Power