
Hyperscale Book: Data Centers Use 20% of Ireland's Electricity and 29% of Oregon City's Water
By Ali Al-Montaser — Senior AI Technical Editor
Journalist Paris Marx's new book Hyperscale reveals that data centers consume over 20% of Ireland's electricity and that Google used 29% of The Dalles, Oregon's water. The book documents how data centers have become a unified political issue across party lines due to their direct impact on electricity bills, water access, and quality of life. Based on a TechCrunch AI interview published November 22, 2026.
Executive Overview
In his new book Hyperscale, investigative journalist Paris Marx documents the immense resource consumption of hyperscale data centers. The book reveals that data centers consume more than 20% of Ireland's electricity and that Google used 29% of the water in The Dalles, Oregon. Based on a rare inside visit to a data center in New Zealand, the book argues that data centers have become a unifying political issue across party lines due to their direct impact on local communities. TechCrunch AI featured an in-depth interview with Marx on November 22, 2026.
📊 Official Data & Technical Specifications Card
| Technical Aspect | Confirmed Official Data |
|---|---|
| 💰 Pricing & Usage Cost | The book and interview do not include direct pricing; the documented economic impact is rising electricity bills in host communities, with Microsoft, Amazon, and OpenAI pledging to fund a larger share of electricity infrastructure. |
| 🌐 Platforms & Immediate Availability | The book is available through the publisher and distributors; media coverage via TechCrunch AI (November 22, 2026). |
| ⚡ Performance & Speed Benchmarks | No model performance benchmarks; documented quantitative indicators: 20%+ of Ireland's electricity, 29% of The Dalles' water, and a moratorium on grid connections around Dublin since 2022. |
| 🛡️ Security & Breach Resistance | Field-documented security measures: passport sent in advance, identity checks upon arrival, card-locked doors in every corridor, surveillance cameras, and perimeter fences. |
| 🧠 Context Window | Not applicable — this is an environmental/political investigation, not an AI model announcement. |
| 🌍 Arabic Language Support & Region | The book does not address Arabic language support; regional impact is not documented in the source, and the phenomenon is global, spanning Ireland, Chile, New Zealand, and the United States. |
Deep-Dive Features & Architecture
Hyperscale offers a rare firsthand account from inside a hyperscale data center in New Zealand, following failed attempts to access facilities in Northern Virginia and Ireland. Marx describes the interior as "utilitarian" rather than cinematic: a warm room where air conditioning vents push air toward servers, with a sharp fan noise that requires earplugs for extended stays. He confirms that security is far higher than expected: a passport sent in advance, identity checks upon arrival, and card-locked doors in every corridor.
The numbers are the core of the book's argument. In Ireland, data centers exceeded 20% of the country's electricity, prompting the grid operator to impose a temporary moratorium on grid connections around Dublin since 2022—before the generative AI boom. In The Dalles, Oregon, where Google built its first owned data centers about 20 years ago, the company sued a journalist from The Oregonian to prevent disclosure of water consumption, before the figure of 29% of the city's water was later revealed. Marx adds the case of Chile around Santiago, where local communities organize protests against water pressures from Google and Microsoft facilities.
Benchmark & Competitive Performance
The source does not include AI model benchmarks, but it provides quantitative comparisons between host communities: Ireland at 20%+ of electricity, The Dalles at 29% of water, versus the grid connection moratorium in Dublin since 2022. Marx notes that these figures predate the generative AI wave, meaning the current cumulative impact is higher than the documented numbers. The author describes the buildout as "inherently unsustainable" and "outside the bounds of a livable planet," arguing that data centers are the "physical manifestation" of technology that can be directly challenged, unlike deleting an app from a phone, which makes little difference to companies.
Industry Impact & Enterprise Adoption
While the book does not directly address the Arab region, its figures translate into tangible operational risks for Arab developers: every large language model called via API consumes a share of data center energy, and rising electricity bills in host communities may become indirect costs for hosting services. The book documents that major tech companies are responding to opposition: Microsoft, Amazon, and OpenAI have pledged to stop using non-disclosure agreements (NDAs) in negotiations with local governments and propose funding a larger share of electricity infrastructure. This shift signals growing pressure on enterprises to account for the environmental footprint of their AI infrastructure and to engage transparently with local stakeholders.
Conclusion
Hyperscale by Paris Marx provides a meticulously documented account of the environmental and social costs of hyperscale data centers, anchored by hard figures: over 20% of Ireland's electricity, 29% of The Dalles' water, and a grid moratorium in Dublin since 2022. The book elevates data centers from a technical footnote to a central political and economic issue, uniting opponents across the political spectrum. For developers, enterprises, and policymakers, it serves as a critical reminder that the digital infrastructure powering generative AI has real-world resource implications that demand transparency and sustainable planning.
Media Source: TechCrunch AI | Fact Verification & Analysis: AI Tools Oasis
Frequently Asked Questions
Data centers in Ireland consume more than 20% of the country's total electricity, prompting the grid operator to impose a temporary moratorium on new data center grid connections around Dublin since 2022.
Google's data centers in The Dalles, Oregon consume 29% of the city's water, according to Paris Marx's book Hyperscale. This figure was revealed after Google sued to prevent disclosure.
The author is journalist Paris Marx. The book investigates the environmental and social impact of hyperscale data centers and generative AI, including electricity and water consumption, noise, and a rare inside visit to a data center in New Zealand.
Because their impact is tangible and local: electricity bills, water access, noise, and quality of life. This has united opponents across the political spectrum, from Ron DeSantis to Bernie Sanders.
Microsoft, Amazon, and OpenAI have pledged to stop using non-disclosure agreements (NDAs) in negotiations with local governments and propose funding a larger share of electricity infrastructure, according to a TechCrunch interview with Paris Marx.

AI Tools Oasis Team
Bringing you the latest news and analysis in the world of Artificial Intelligence with accuracy and credibility. Follow us for all updates.


