Brief Author Note

Life was lifing this past weekend! That’s why you’re getting this newsletter issue at 7PM, instead of 7AM in Cambodia time lol.

I was on a podcast episode recently with Aashka Patel and Scot Bryson discussing the intersection of AI, sustainability, and water resources.

There’s a special part about reusing wasted heat from data centers in order to grow tomatoes in Canada through Scot’s work.

Enjoy listening!

Chau,

Nate

Quick Links!

• Podcast Episodes

• Scientific Papers

• Policy Documents

• Multimedia

• Organizations

• Tools

• Fellowships Corner

• Submit Resources

Episode 8: The Elephants on the Horizon (AI & Policy/ESG): This Legal Sustainability Alliance episode explores the growing sustainability risks on the legal sector’s horizon, including AI-driven environmental impacts and regulatory shifts. Features insights from sector leaders on adapting legal services for environmental responsibility.

Decoding the Planet: From Whales to Whistleblowers (AI & Environment): Mozilla’s IRL podcast dives into the digital carbon footprint of online content, examining how deleted data impacts climate goals. Guests incuding Dr. Sasha Luccinoi discuss AI’s role in optimizing data storage, energy use, and shaping sustainable digital behaviors.

AI for Impact: Scaling Circular Economy and Sustainable Supply Chains (AI & Social/Economic Impacts): Sushma Kittali Weidner shares how AI technologies are transforming circular economy strategies and improving supply chain sustainability. The episode highlights innovations, metrics, and real-world case studies where AI supports ESG-aligned goals.

Thermodynamic Computing System for AI Applications (AI & Data Centers/Hardware): This Nature Communications article introduces a novel physics-based hardware component including the stochastic processing unit, built with RLC circuit cells that leverages thermodynamic computing to perform AI primitives such as Gaussian sampling and matrix inversion. It offers a promising low-power, high-speed alternative to GPUs for probabilistic AI tasks.

Evaluating the Energy-Efficiency of the Code Generated by LLMs (AI & Data Centers/Hardware): This arXiv preprint analyzes energy and performance metrics of code generated by 20 popular LLMs across 878 LeetCode problems. The study reveals that LLM-generated code often lags behind human-written solutions in efficiency, consuming significantly more energy—sometimes up to 450× for complex tasks. Models like DeepSeek‑v3 and GPT‑4o perform best among the LLMs tested, while others like Grok‑2 and Gemini‑1.5‑Pro are far less efficient.

Advancing Hourly Gross Primary Productivity Mapping Over East Asia Using LGBM (AI & Climate Science): This study presents a new machine learning (Light Gradient Boosting Machine) approach to generate high-resolution, hourly maps of gross primary productivity (GPP) for East Asia (2020–2021). The paper highlights clear diurnal patterns across different land cover types and latitudes, offering valuable insights for ecosystem monitoring and carbon cycle modeling.

Big Tech’s Climate Performance and Policy Implications for the UK (AI & Sustainability): This report critically examines the carbon disclosures and climate commitments of major technology firms operating in the UK. It evaluates the alignment of corporate practices with net zero goals and proposes a framework for national policy interventions to regulate the climate impacts of digital infrastructure and AI deployment.

Powering North America’s Data Center Boom (AI & Infrastructure): This white paper outlines the projected energy demands of North America's rapidly expanding data center ecosystem. It highlights AI workloads as a major driver of growth and advocates for grid-aware design, clean energy integration, and smarter infrastructure planning to mitigate environmental strain.

AI for Climate Action in Developing Countries: Opportunities, Challenges and Risks (AI & Climate): This technical overview explores the role of AI in supporting climate action in developing countries, focusing on applications like early warning systems, agriculture, and emissions monitoring. It addresses the digital divide, ethical considerations, and the need for inclusive governance to ensure equitable AI deployment under the Paris Agreement.

The Slow Death of Scaling and What Comes Next (AI & Climate Science/Research): In this keynote from Khipu 2025 in Santiago, Chile, AI researcher Sara Hooker examines the limitations of current scale-driven deep learning architectures and explores alternative approaches such as modular and thermodynamic models that could lead next-generation, climate-aware AI.

Saving the Planet with AI Keynote Speech (AI & Environmental Impact): Watch the keynote speech of Daniel Erasmus, CEO, ClimateGPT / Full Member Club of Rome about saving the planet potentially with AI.

Episode 5 - AI/ML are flipping Traditional Data-to-Theory Method to Accelerate Climate Change Understanding (AI & Climate Science/Research): Watch this lecture of how AI and ML are flipping traditional data-to-theroy methods to accelerate our understanding of climate change systems.

DejaBlue (AI & Energy): DejaBlue supports businesses in deploying and managing large-scale electric vehicle charging and smart energy management onsite to accelerate the transition towards sustainable mobility.

Enter (AI & Infrastructure): It's AI-powered platform for energy-efficient building renovations are used across Germany.

Underheat (AI & Infrastructure): We use AI to design and manufacture more efficient heating systems for heat pumps, reducing the cost and disruption of retrofit upgrades.

AQUA-7B (AI & Ocean Conservation): AQUA-7B is Kurma AI’s flagship 7-billion parameter large language model built exclusively for the global aquaculture industry. And it is the first large language model for the aquaculture. It is fine-tuned to deliver actionable insights for aquaculture species-specific farming, hatchery operations, water quality control, and disease management.

Environmental Monitoring System (AI & Biodiversity/Ecology): Arduino and sensor-based system for monitoring temperature, humidity, and air quality at local scale.

FABRIC: When Servers Meet Species (AI & Biodiversity/Ecology): FABRIC (Fab-to-grave Analysis of Biodiversity-Related Impact of Computing) is a comprehensive framework for assessing the environmental impacts of computing hardware throughout their entire lifecycle, with a particular focus on biodiversity-related impact categories with the code.

Klarna AI Climate Resilience Fellowship (Open to global applications): The AI for Climate Resilience Program is a new initiative by Klarna that aims to support pioneering projects that leverage artificial intelligence for climate adaptation in underserved, climate-vulnerable regions. 🗓️ Deadline: August 31, 2025.

The Boring Fund (Open to global applications): Thanks to the support of Arm, the company building the future of computing, The Boring Fund is now providing $80,000 to fund the "boring" but necessary work that makes conservation technology more accessible, sustainable, and effective. If you have an idea that strengthens the foundations of this field, apply today! 🗓️ Deadline: August 19, 2025.

Schmit AI in Science Postdoctoral Fellowship (UK): The Eric and Wendy Schmidt AI in Science Postdoctoral Fellowship, a programme of Schmidt Futures, enables researchers in STEM to apply Artificial Intelligence (AI) and Machine Learning (ML) techniques in innovative ways. The scheme supports research that integrates AI/ML with internationally leading science, with a goal of achieving step-changes in understanding or application. 🗓️ Deadline: August 12, 2025.

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. 🙃 

That’s it for this week.

Thanks for reading the 8th issue of The Climate Code! It means a lot to me and we have more coming in the future, so definitely stick around! 🏃‍♂️💛

Nate

P.S.

New here? Check out the AI & Environment Resource Hub if you have no idea what it is. 😆