Energy Projects Supporting AI Growth

AI Doesn’t Float! Data Centers are anchored in physical infrastructure with real-world trade-offs.   In our collective pursuit of AI acceleration, one question often goes unasked: Where will all the power come from and who will bear the cost of delivering it? While headlines warn that data center emissions may hit 3–4% of global CO₂ by 2030, the IEA ( https://lnkd.in/eKe89HjZ) estimates this number closer to 1%. So why the wide gap? Because location, timing, and grid readiness matter more than averages. In places like Dublin, data centers already consume nearly 20% of available power. In the U.S., data centers may surpass the electricity demand of all domestic heavy industry combined by 2030. This is a planning crisis already underway. Through my recent advisory work, I’ve seen how forward-thinking infrastructure planning, decarbonization strategy, and locational modeling can either unlock or bottleneck entire regions. Data center clusters need grid flexibility, clean firm power, and local engagement. However, very often we see a race to build without a strategy to share or sustain.   The IEA’s new report (“Energy and AI” https://lnkd.in/eKe89HjZ) lays out three imperatives: 1. Diversify clean power supply (yes, including geothermal, SMRs, and batteries). 2. Accelerate grid build-out, not just generation. 3. Foster better collaboration between tech and energy planners, upstream, not after permits are filed.   A sustainable AI future is about cross stakeholder coordination not solely carbon emissions. Cities, utilities, and communities deserve a seat at the table before another megawatt is claimed.     #EnergyTransition #DataCenters #GridPlanning #SustainableAI #IEA #Infrastructure #ClimateTech #Resilience

⚡ AI Needs Power. Renewables Are Faster, Cheaper, and Smarter. ⚡ As generative AI explodes, so does electricity demand. But Trump’s “AI master plan” skips over a basic truth: You can’t power the future with yesterday’s infrastructure. What the plan ignores: 🔋 Speed to power is everything. Data centers need gigawatts of new capacity—now. • Solar + storage projects are coming online in under 24 months. • New natural gas plants can take 5–7 years, plus years of permitting delays. 📉 Clean energy beats gas on cost. • In 2024, the average PPA price for utility solar+storage in the U.S. was $45/MWh, while new gas peakers often exceed $90–120/MWh. • Texas is a case in point: Wind and solar saved ERCOT customers $11B in 2023 alone. 🔥 Natural gas is slow, expensive, and volatile. • Gas prices surged 700% in Europe during the 2022 crisis. • In California, battery storage now covers peak evening demand formerly met by gas—at lower cost and with more grid flexibility. 💡 Smart AI needs smart energy. A credible AI strategy should accelerate investment in clean, fast, modular energy—not double down on fossil bottlenecks. Instead of propping up gas, we should be scaling: ✅ Solar + wind ✅ Grid-scale batteries ✅ Virtual power plants ✅ Clean, responsive demand AI needs a power plan. Let’s make it 21st century.

AI will now help build nuclear reactors 🔗 Westinghouse and Google Cloud have teamed up to bring artificial intelligence into the heart of nuclear energy. Their goal? Make reactor construction faster, smarter, and easier to replicate, and bring more data-driven insights into existing nuclear plant operations. 🧠 At the center of this effort are Westinghouse’s AI platforms, HiVE™ and bertha™, now integrated with Google Cloud tools like Vertex AI, Gemini, and BigQuery. Using these platforms together, the two companies have already pulled off a key milestone: autonomously generating optimized modular construction work packages for the AP1000 reactor using AI. That means fewer delays, more standardization, and a real step toward scalable nuclear builds. ⚙️But it doesn’t stop at construction. These AI tools will also help operate reactors more efficiently, improving diagnostics, maintenance scheduling, and long-term performance. ⚡This partnership lands at a critical moment. Data centers, driven by AI growth, are on track to double electricity demand in the coming decade. That demand can’t be met with fragile grids or weather-dependent sources alone. We’ll need firm, clean energy that runs 24/7, and that’s exactly what nuclear can provide. 🌍 So yes, Google is investing in the future of energy, not just computing. By helping optimize new reactor deployment, it’s laying the foundation for a grid that can keep up with the very technologies it’s advancing. 🔋AI won’t just consume power. It can help build the systems that generate it. Source: https://lnkd.in/d4YQUsun #NuclearEnergy #AI #GoogleCloud #Westinghouse #CleanEnergy #EnergyInnovation #AP1000 #eVinci #AP300 #HiVE #bertha #AIinEnergy #DigitalTwin #VertexAI #BigQuery #FutureOfEnergy #AdvancedNuclear #ModularConstruction

Two major technology transformations came into alignment today in Washington, DC: AI and fusion energy. At an artificial intelligence event I attended today, the White House released an AI action plan that calls for the US to put a priority on reliable, dispatchable power and specifically to “embrace new energy generation sources at the technological frontier,” including fusion energy. Here’s why this is important. AI is a major technological force, but its data centers need electricity. OpenAI, the maker of the ChatGPT AI tool, said this week it expects to have more than a total of 1 million AI chips in service by the end of 2025 — and after that they’re headed toward 100 million. That’s a tremendous amount of new power demand. When we put fusion on the power grid in the early 2030s, we think it’ll be the ideal power source not just for data centers but to help power the global economy’s energy needs. Fusion power plants can be built just about anywhere, with essentially unlimited fuel and no emissions. Fusion has no reliance on pipelines or weather, and it’ll operate 24x7. Google, another major AI force with its Gemini tool, is helping us to build this fusion future with a new investment in CFS and an agreement to buy 200 megawatts of electricity from our first ARC power plant. That partnership, which we announced in June, is a clear sign there’s demand for what we’re offering. As I told Bloomberg News this morning about our Google deal, “Hyperscalers are fundamentally techno-optimists that believe in a future that is abundant and the technology can be used to solve problems and enable new things. Having that involved with fusion is a really good match. That's a big signal that fusion is wanted.” The government can help bring fusion’s benefits to the grid faster. Fully funding programs like the U.S. Department of Energy (DOE)’s Milestone-Based Fusion Development Program would help several fusion energy companies, including CFS, so people and businesses can benefit from this new source of energy as soon as possible. Scroll to the comments to catch the link to the Bloomberg TV interview. #FusionEnergy

Data centers that run the Internet & AI computations will face increasing political and market pressures as prices & emissions rise. A way forward is to invest in communities with enough clean power to share, invest in infrastructure to keep costs down, speed clean energy connections, & get to zero water consumption. The Department of Energy's non-profit foundation, the The Foundation for Energy Security and Innovation (FESI) could help launch Smart AI Fast Lanes. My Policy Brief for Federation of American Scientists outlines some policy pathways: https://lnkd.in/eYchGvZh

Earlier this week, Anthropic released their report on energy infrastructure needed to support AI data center build out in the US. Lots of interesting facts, estimates, and ideas in this report: "Leading AI developers and other experts expect that training a single state-of-the-art AI model will require a 5GW data center by 2028-2030. That trajectory aligns with our own projections at Anthropic. If the United States unlocks its energy sector’s potential, we anticipate using 2GW and 5GW data centers for a single AI training run in 2027 and 2028, respectively. With a handful of U.S. companies competing at the AI frontier and in need of similar capacity for their most advanced training runs, that statistic implies 20-25GW required in total for frontier AI training by 2028, split across several locations. And many expect that in the coming years inference will use roughly as much or more compute and energy compared to training." The report makes detailed recommendations on how to facilitate the development of the needed energy infrastructure including generation, transmission, interconnection, equipment, and permitting. https://lnkd.in/gGhBrDKD