Environmental Impact Of Technology

A big question looms over generative AI: what really is its impact on the environment? I spent months investigating a single campus of Microsoft data centers in the Arizona desert - designated in part for OpenAI - in an attempt to find out. The process underscored just how little visibility we have into basic information, such as the water and energy consumption of these silicon monstrosities, which are now being built at an unprecedented rate, including in the desert. While Microsoft has invested massively to improve the sustainability of its data centers, it is also a for-profit company. At times it has suppressed environmental impact measures or pushed the opposite narrative from internal projections, even as employees urged more transparency. Meanwhile, after I FOIA'ed several agencies at the state, county, and city levels, the city returned relevant docs with all of the numbers redacted. (Screenshot attached.) Neither are willing to inform the public about the real-world costs supporting this technological wave amid an accelerating global climate crisis. My latest for The Atlantic. https://lnkd.in/guPJs8wZ

A bold prediction no one wants to hear: Half of all commercial solar systems installed before 2016 will be underperforming or non-operational by 2030. The solar industry is obsessed with the future. Cutting-edge panels (bigger is better). Sleek batteries. Dazzling projections for new installs. But here's the reality we can't afford to ignore: a silent crisis unfolding on rooftops across America—a crisis I've been tackling firsthand since 2012, traveling the country with SunPower to address some of the industry’s most pressing system failures. Across the country, tens of thousands of rooftop solar systems—once hailed as the clean energy revolution—are quietly decaying. Not because the technology failed, but because the industry did. We rushed to install. We cut corners. We promised 25 years of performance… and delivered systems that can’t make it past 10. Here’s what’s killing them: Inverters are dying—many are already out of warranty, with no replacements available. Wiring and electrical infrastructure that was never designed for 25+ years of exposure. Install quality? Forget it—an army of barely trained crews built the boom, and now we’re paying the price. Maintenance? There was no plan. Just a contract, a handshake, and a hope it would all work out. This is not just an engineering issue—it's a financial one. Underperforming assets are generating less revenue than forecasted, while increasing the risk of electrical faults, fire hazards, and insurance claims. And here's the kicker: almost no one is ready to deal with this wave of system failures. Asset managers, facility owners, and even EPCs are discovering that repowering, remediation, or decommissioning is far more complex and expensive than expected. This is where the next frontier of solar energy lies—not in installing the next 100GW—it’s rescuing the first 100GW. Revitalization. Repowering. Responsible end-of-life planning. The question isn’t whether it’s coming. It’s whether we have the guts to face it. Are we going to keep pitching the dream— —or finally clean up the mess we left behind?

Fossil-free heat for industry has long been one of the toughest challenges in the energy transition — accounting for a significant share of global energy demand and emissions. That’s why the latest sand-battery developments are so exciting. A Finnish cleantech startup has now deployed an industrial-scale sand-based thermal battery that converts renewable electricity into high-temperature heat stored in sand — and delivers it on-demand as steam for industrial processes.

People don’t pay for green. Full stop. We see many #climatetech startups marketing their products in this order: 1️⃣ Sustainability - the products are green and have low carbon intensity. 2️⃣ Resilient supply chain - the sourcing of the product is done in a more resilient and reliable way. 3️⃣ Performance - the product is (or nearly is) a drop-in solution. 4️⃣ Price - there is currently a “green premium,” but it will decrease as we scale. Yet, time and again, these companies, especially those selling commodities, experience pushback from an industry unwilling to buy these goods and narratives. The reason is that the industry has the exact opposite set of priorities: 1️⃣ Price - in a high-interest environment where margins are eroded and many businesses face fierce competition (e.g., from China), price parity is the top priority. Even a few cents per kW/h or gallon can make a difference. I recently learned of a battery startup whose raw materials alone cost more than the fully assembled battery of a Chinese competitor. No one will pay that premium. 2️⃣ Performance - many new solutions promise technical performance improvements, but most are not packaged to qualify for all customer requirements and have little evidence to prove long-term benefits. In mega projects, durability is almost always more important than unproven superior performance. Sunfire is flourishing because of their Alkaline cells, not their SoX full cells. 3️⃣ Resilience - following the pandemic and the scarcity of raw materials, this is indeed a growing concern for both industry and governments. 4️⃣ Sustainability - if a product can address all the above topics and also be green, the industry will be happy to adopt it. What does this mean? Startups need to take a market-centric rather than a tech-centric approach. They should develop their go-to-market strategy from day 1 to prioritise customers whose needs align most with their story, and design their entire product and value proposition around those customers requirements. For example, a raw material startup shouldn’t target the battery industry where price and quality are crucial. Instead, they might find success selling to the cement industry, where quality is less critical, and there’s a whole new value proposition around cirularity and sustainability. #venturecapital #fundraising #productmarketfit

Solar Panel Waste is Tiny—Coal & Gas Emit Hundreds Of Times Mass Per MWh Every few months, the same tired narrative resurfaces: solar waste is the next environmental crisis. It’s not. It’s a distraction. Full article: https://lnkd.in/gtkB5pfM Yes, solar panel waste exists—nearly 900,000 tonnes in 2024—but the vast majority so far is from early retirements, not panels reaching end-of-life. Most are storm-damaged, replaced for newer tech, or reflect early manufacturing flaws. That’s normal for maturing technologies. But here’s the kicker: even with early replacements, solar produces just ~2 kg of solid waste per MWh over 25 years. Compare that to coal’s ~90 kg of toxic ash plus nearly a tonne of CO₂ per MWh. Gas? Around 450 kg of CO₂ per MWh, no solid waste—but plenty of invisible damage. Solar panel materials—glass, aluminum, silicon—are mostly recyclable and benign. Coal ash is laced with arsenic, mercury, and uranium. One contaminates soil and kills ecosystems. The other sits quietly in recycling yards. Let’s be blunt. The “solar waste crisis” is fossil fuel PR spin. The real crisis is 100+ billion tonnes of CO₂ dumped into the atmosphere over five years by coal and gas. That’s the catastrophe. That’s what’s damaging the environment, not PV panels. We have scalable solar recycling systems. What we don’t have is a way to suck coal ash out of the groundwater or CO₂ out of the sky fast enough to matter. This is very similar to the tired narrative about wind turbine blades in landfills when coal emits 3,000 times the mass of CO2 per MWh, and gas 1,500 times. Solar panel and wind turbine end-of-life waste isn’t the crisis. Fossil fuel waste is.

We’ve been treating AI’s energy problem as a datacenter problem. But the datacenter is the most optimized part of the stack. The real inefficiency is happening higher up. Google estimates that 60% of AI’s energy now comes from inference. Meta says 60 to 70%. AWS, 80-90% of its ML compute demand. A single prompt is insignificant. Billions across apps, agents & API calls aren’t. We’re on track for trillions. The mismatch? We’re optimizing infrastructure while most of the energy and cost are created at the application layer. Even the cleanest datacenter can’t compensate for a stack that sends every request to a frontier model, runs jobs at the costliest hours of the grid, and treats all queries as equally urgent. The biggest gains won’t come from better cooling or more renewable PPAs. They will come from how we design, route & operate the models themselves. There is a way to architect sustainability as a first-order principle in the AI lifecycle itself. So what does a more efficient stack look like? I’m seeing some really cool stuff these days. It begins with grid-aware infra. Platforms like Emerald AI align compute with renewables, shifting batch workloads to cleaner hours and routing traffic to cleaner regions. Crusoe rethinks the foundation entirely by converting stranded natural gas and heat recovery into compute. Training visibility changes how teams build. CodeCarbon exposes the emissions of every experiment forcing real decisions. Once numbers are visible, priorities shift. Does a 2% accuracy gain justify 10× more compute? Then comes inference intelligence. ChatGPT's routing prevents unnecessary over-computing, while GreenPT builds efficiency into the foundation so every inference run uses less power by default. One optimizes after building. The other designs for efficiency from the start. This is where FinOps & sustainability converge. User visibility matters too. Most people have no idea how much energy their prompts consume. When that information becomes visible in real time, behavior shifts. People choose lighter models, batch calls, and refine their prompting. Shared baselines are emerging. The GSF’s SCI turns sustainability into a measurable standard. GPU-level tools like Neuralwatt replace estimates with real power data and expose waste at the hardware level. But none of this works if the layers stay disconnected. This is the logic behind Antarctica: a single observability layer that connects cost, usage, energy, and user behavior across cloud and AI. Grid carbon intensity, training emissions, inference energy, hardware telemetry, and application analytics converge into one source of truth. To make inefficiency measurable at the point of decision. And in AI, every inefficiency appears twice: once as wasted energy and once as wasted dollars. So let’s make this practical now. I’m putting together a shared list of tools that actually improve efficiency across the AI stack. Which ones would you recommend?

Just read fascinating research on AI energy consumption - but the methodology is almost more interesting than the results. The Challenge: How do you measure the carbon footprint of Claude, GPT-4, or any commercial AI when the companies won't share their infrastructure data? The Solution: Researchers combined public API performance data with reverse engineering: ⌚️ Step 1: Scraped latency and tokens-per-second from artificialanalysis(dot)ai across 30 models 📈 Step 2: Used statistical inference to estimate hardware (Claude likely runs on AWS H100/H200 based on performance patterns) ⚡️ Step 3: Applied the formula: Energy = (inference time) × (estimated GPU power + system overhead) × datacenter efficiency 💦 Step 4: Layered in region-specific multipliers for carbon intensity and water usage They didn't need Anthropic's internal data. They used publicly observable performance to work backward to energy consumption. Key Finding: Claude-3.7 Sonnet scored highest (0.886) in eco-efficiency, while DeepSeek models used 70x more energy than GPT-4.1 nano. When companies won't publish sustainability metrics, researchers find creative ways to find out anyway. It's the core philosophy of the Impact Framework, if you can observe something, you can measure it's impact. Even if an organizations is not disclosing, they might be leaking enough observable information that you can model the impacts anyway. Transparency through ingenuity. 👏 Great work Nidhal Jegham, Marwan F. Abdelatti, El Moubarki Lassaad and Hend Abdeltawab

Today, I came across a 'Sustainability Compliance Tech Map' designed to guide companies through the maze of solutions for compliance. It is a helpful resource, but it raises a critical question: Are we going to create new tools for each regulation? Or can we embed compliance into the digital backbone we’re already building? Reflecting on the Budapest Declaration, I encourage our industry to rethink this approach. Rather than inventing new solutions, let us leverage what we already have. At 9altitudes, we base our projects on powerful platforms like Microsoft, PTC, and Tulip Interfaces - solutions that support robust, scalable digital common threads across industries and enable integrated, data-driven compliance. Compliance should not be a standalone task. It should be a natural extension of the CAD, PLM, MES, ERP, and commerce platforms we rely on daily. By embedding compliance into these systems, it becomes an integral part of operations, connecting all data seamlessly. Industry efforts like the IDTA - Industrial Digital Twin Association, using frameworks like the Asset Administration Shell (AAS), already bridge systems for initiatives like the Digital Product Passport (DPP) without adding unnecessary complexity. Similarly, Microsoft Purview Compliance Manager helps companies assess and manage compliance across multicloud environments, building on existing architectures rather than creating silos. The future of compliance lies in enhancing our digital thread with smart data layers that integrate, communicate, and govern information across functions. Let us use this moment as a call to action. Compliance should not be a burden but a seamless part of the journey - helping us build a sustainable, resilient ecosystem for the future. I invite our colleagues, partners, and industry leaders to share their perspectives. Are you using existing platforms or adding new layers? Let us discuss how we can collectively build a sustainable future by leveraging the solutions we already have. Please feel free to comment, share, or engage with your thoughts. Together, we can make compliance smarter, simpler, and truly impactful. With all respect: what we need is not more legislation or more tech maps - it is a commitment to maximizing the solutions we already have, leveraging them to build a sustainable future. Agree ? #DigitalThread #Sustainability #Compliance #BudapestDeclaration #Microsoft #PTC #Tulip #9altitudes #Industry40 #Industry50 #DigitalTwin #ERP #PLM #MES

The power of innovation to drive change never ceases to amaze me. We’ve announced the selected 11 groundbreaking startups for our Amazon Sustainability Accelerator's ClimateTech cohort. From an impressive pool of over 550 applicants, these companies stood out for their bold approach to environmental challenges.   Let me share three innovations that particularly inspired me: - Shayp (Belgium) Their AI-powered water monitoring system could revolutionize how buildings manage water consumption, potentially reducing usage by 20%. This is the kind of practical innovation that can create immediate impact. - NANDO (Italy) As an Italian, I'm particularly proud to see this team's AI-powered waste management system. Using smart cameras for real-time waste analysis, they're transforming how we think about recycling efficiency. - Mhor Energy Limited (UK) Their liquid-form battery storage solution is exactly the kind of thinking we need - practical, efficient, and designed for real-world conditions.   The potential of impact at scale is impressive. Last year alone, we invested €750,000 in pilot programs from our previous cohort, and I'm eager to see how these new technologies could transform our European operations. I'd love to hear your thoughts: what sustainability innovation you’re most excited by? https://lnkd.in/dcqM97kv

You think Silicon Valley is the future of climate tech? You couldn’t be more wrong... The most meaningful progress is happening far from the venture bubble, in small labs, research stations, and community workshops where the focus is on solving practical problems rather than chasing scale. 2025 has been a record year for climate tech investment. But the real story isn’t how much money is being raised. It’s what that money is building. The direction of innovation is shifting toward systems that are modular, verifiable, and built for real-world conditions. These technologies can be deployed quickly, maintained locally, and adapted to places that can’t wait for large infrastructure to arrive. 🌱 Releaf Earth (YC 2025) converts food waste into biochar that restores soil, locks carbon, and produces renewable power for local microgrids. Their portable reactors make it possible for small communities to build their own carbon markets. Biochar now accounts for more than 90 percent of all durable carbon removals delivered globally, showing how central this technology has become to practical decarbonization. 🌱 Modular Green Hydrogen startups in programs such as RMI’s accelerator are proving that hydrogen production doesn’t have to rely on billion-dollar plants. Their systems use renewables and recycled water to power rural transport and small industries, aligning closely with the U.S. 45Q incentive for low-carbon hydrogen. 🌱 Recyclable wind turbines built from bio-resins and nanocellulose are beginning to close the loop on renewable energy. They address a long-standing issue in the sector, how to manage the waste created when turbine blades reach the end of their life. 🌱 Bamboo-based cooling panels, now emerging from university and startup labs, use natural condensation to lower indoor temperatures without electricity. Early trials in Asia and Africa suggest they could offer low-cost cooling in regions already struggling with extreme heat and limited access to power. 🌱 AI and satellite mapping tools from companies such as Astraea are providing live, high-resolution data on climate risks. What used to take months of modeling can now be updated continuously, helping governments, insurers, and local planners make faster, better decisions. These examples point to a wider shift. Climate technology is no longer defined by size or spectacle. It is defined by systems that are reliable, measurable, and designed for real contexts. Policies like the European Union’s Carbon Removal Certification Framework are reinforcing this trend, directing investment toward solutions that can demonstrate genuine and lasting impact. The next phase of climate innovation will not be driven by how much it raises or how fast it scales. It will be judged by how well it works, consistently, locally, and over time.