WELL… Turns out the obvious was stated after the four tech companies released their Quarter 2 (Q2) 2026 reports from last week: the CapEx for AI data centers is too damn high! It keeps going up and up every quarter lol. At this rate, we’re really going to be hitting the trillion dollar arena by 2027 with these tech companies shelling out mola mola for AI data center infrastructure in the U.S. 💵

Check out these numbers that I found after going through the Q2 2026 documents from Amazon, Alphabet/Google, Microsoft, and Meta (check out my meme at the bottom of the issue):

Winners:
1️⃣ Amazon - $54.21B
2️⃣ Alphabet/Google - $44.92B
3️⃣ Microsoft - $35.80B
4️⃣ Meta - $31.08B

Seems like we’re going to be ramping out throughout the rest of the year, prime those engines folks! 🚂

Chau 👋,

Nate

The Truth About AI Data Centers, Power & America's Future with Travis Hawkes & Eric Sonner (AI & Data Centers/Hardware): In this episode of Ever Onward, Tommy Ahlquist and Travis Hawkes sit down with Eric Sonner, founder and CEO of Data Airflow, to explore one of the fastest-growing industries in the world: AI data centers.

Jennifer Granholm's Data Center Playbook for Governors (AI & Data Centers/Hardware): Jennifer Granholm has held both roles, as governor of Michigan and as Secretary of Energy. This week on Energy Empire she laid out how a governor should handle the data center boom, and it was the sharpest version of that playbook I have heard.

Nobody Cares About Sustainability Anymore? Let's Find Out with Electricity Maps (AI & Energy): In today's episode, we're joined by Dragos, Marketing Lead at Electricity Maps, the platform building the world's most comprehensive real-time view of the electricity grid, showing where every country's power actually comes from.

Flexibility-Aware Framework for Efficient Planner-Initiated Siting of Data Center (AI & Data Centers/Hardware/Model): Explosive growth in energy-intensive AI data centers is outstripping the pace of power grid interconnection and transmission expansion. This work introduces a planner-initiated siting framework that combines reliability-gated screening, system-wide market-impact assessment, and entropy-weighted scoring to produce ranked, pre-certified catalogues of interconnection-ready locations. Authors: Kim et al. Journal: Nature Communications.

Looking to the Brain to Improve Energy Efficiency of AI (AI & Data Centers/Hardware/Model): Modern AI systems have achieved remarkable capabilities, but at an extraordinary energy cost. This review identifies key biological principles that support energy-efficient capacities in biological brains, and considers how they might inform the design of more sustainable artificial systems. Authors: Peters et al. Journal: Current Biology.

NetCarbTrace: A Probing Tool to Measure and Explore the Carbon Footprint of Computer Networks (AI & Data Centers/Hardware/Model): Current methods for calculating sustainability metrics of network links rely on fixed-rate estimations. This paper presents a context-aware model of energy consumption, along with an open-source tool that enables developers to use the Internet more efficiently. Authors: Vergallo et al. Journal: IEEE Software.

Sustainability in the Nordic Data Centers (AI & Infrastructure): A look at how data center operations in the Nordics are approaching sustainability, from energy sourcing to regional infrastructure. Authoring Body: Digital Realty.

Sustainable AI Agent Design: Building Intelligence We're Willing to Live With (AI & Sustainability): A framework for designing AI agents that balance capability with long-term sustainability and responsible resource use. Authoring Body: N-Able.

Principles for Sustainable Water Use by Data Centers: Building More Effective Public-Private Collaboration (AI & Infrastructure): A set of guiding principles for sustainable water use by data centers, with an emphasis on stronger public-private collaboration. Authoring Body: Water AI Nexus.

Powering the Data Center Boom Responsibly (AI & Energy): Data centers power artificial intelligence, digital services, and much of the modern economy. But their rapid expansion across the United States is driving steep growth in electricity demand and putting pressure on water, land, and local communities.

Green AI: Making Machine Learning Environmentally Sustainable (AI & Environmental Impact): After considering the significance of the carbon footprint of AI, Charles offers practical strategies to reduce environmental impact at each stage of the AI lifecycle, including smaller datasets, transfer learning, model compression, and edge computing.

Illuminating a More Sustainable Future | Conversations on Sustainability (AI & Environmental Impact): A conversation with Szymon Slupik, CTO and co-founder of Silvair, on why sensors are key to a more sustainable future, and how connected lighting networks optimize building operations, reduce waste, and extend the life of critical infrastructure.

FuelCell Energy (AI & Energy): FuelCell Energy is an American clean energy company delivering continuous, scalable power to support mission critical applications and grid resilience. Location: North America.

ECL (AI & Data Centers/Hardware): The world's first fully-sustainable, hydrogen-powered, off-grid data center-as-a-service. Location: Other.

Calibrant Energy (AI & Energy): Calibrant is a leading provider of on-site energy solutions for large power users. Location: North America.

Data Center Air Pollution Tracker (AI & Pollution Response): This scorecard evaluates the eight hyperscaler tech and AI companies with U.S.-based data centers on one simple question: are their data centers running on clean energy, or fossil fuels like gas, oil, and coal? Price: FREE.

Legal OSS (AI & Workforce Development): LegalOSS is a small index of open-source legal software. Every entry is a real GitHub repository, its stats come straight from GitHub, and each repository can be listed exactly once. Browsing needs no account. Price: FREE.

Berkeley Decarbonization Tool (AI & Sustainability): Designed to bring consistency, transparency, and scientific robustness to HVAC decarbonization studies, reducing reliance on proprietary or ad hoc internal modeling approaches. Price: FREE.

If you have resources that you wish to submit to the AI & Environment Resource Hub, fill out the Google Form! Thank you for suggesting resources to add to the Resource Hub. 🙃

Yours truly created this meme after looking at the Q2 2026 total capex amounts that I could find with Amazon, Alphabet/Google, Microsoft, and Meta lol.

Thanks for reading the 52nd issue of The Climate Code! It means a lot to me and we have more coming in the future, so definitely stick around! 💚

Nate

PS 😁

New here? I would recommend checking out Data Center Battlegrounds (our YouTube channel focusing on the environmental impacts of data centers) and also the AI & Environment Resource Hub itself.

PPS 😆

If you’re feeling EXTRA GUTSY, connect with me here on LinkedIn! Say hey 👋.