When Energy and Intelligence Converged 2025-12-11 · 13 min · Human Pace, The Audio Edition ---------------------------------------------------------------- Hi, I'm T. I translate risk for people making capital decisions in clean energy and artificial intelligence infrastructure. This is Human Pace, the audio edition. Today's piece is called Year End Review - 2025 When Energy and Intelligence Converged. For me, 2025 will likely be remembered as the year clean energy turbulence met the massive arrival of artificial intelligence. The timing felt odd. Maybe even unfortunate. Yet the collision revealed something essential about progress. Progress rarely moves in a straight line. It still has a direction. Clean energy navigated a year of constriction and recalibration. Projects slowed. Capital tightened. Supply chains reminded us of their dependencies — and onshoring, nearshoring, and friendshoring take time. The scale of compute demanded new power. The scale of risk demanded new thinking. Together, these shifts left me with three thoughts. The first is about rebalancing. 2025’s volatility was not a failure of the energy transition — it was the transition working, correcting, adapting, and readjusting to find equilibrium. Rebalancing isn’t an annual event on a schedule. It simply happens — and it requires us to maintain perspective through volatility. The second is about pace and judgment. Move too quickly, and the system becomes fragile. Move too slowly, and good judgment arrives after the decision has already been made. Acceleration is powerful, but direction is leadership. And what kind of leadership? Energy leaders who understand AI. AI leaders who understand climate. Business leaders who understand all three. The third thought is the one I carried forward most strongly: the future belongs to people who can see across domains. You may have read different numbers on this and the statistics shift depending on when you look and who is counting. But they all point the same direction. In 2025, China installed more solar and wind in a single year than the entire renewable capacity of the United States. That scale is staggering. Scale matters and the coherence matters more. So, the next chapter of the energy transition belongs to those who can align energy, climate, artificial intelligence, and infrastructure into something resilient and investable. Because climate change doesn’t pause, it compounds. If there was a whisper running underneath 2025, it was this: progress is built, and progress has to be stewarded. Let's celebrate what was built. Let's learn from what did not work. Let's carry both into the next one. Which raises a question I keep coming back to. Why are we here at all? If the climate were mild and stable, would we obsess about which energy source we use? We set out to solve energy in order to solve climate. We solve climate so the humans can live and thrive on planet Earth. AI or no AI — that mission is unchanged. So, as we step into 2026. May our decisions be shaped by clarity, not just projected momentum. By courage, not the convenience of chasing whatever is hot this quarter. And by the quiet confidence that comes from understanding the systems that power a changing world. Those are my reflections. Next, ten charts accompany the written piece. If you have the written version open, they are in this order. But if you don't, I'll describe the shape of each one and why it matters. Start with the atmosphere. Figure one. Global greenhouse gas emissions, 2019 through 2023, from the Rhodium Group. Their line starts at 51.1 gigatons of carbon dioxide equivalent in 2019. It drops to 48.7 in 2020, when the pandemic shut the world down. From there it climbs back — 50.8, 51.2, and finally 51.8 by 2023. The dip was real. It was also temporary. Five years and one global shutdown later, the world is emitting more than it did before. That is the baseline every other chart here has to work against. Now follow the money. Figure two. Where the renewable capital actually went in 2024. The International Renewable Energy Agency — IRENA — maps 807 billion dollars of renewable energy investment across the globe, up from 662 billion in the two years prior. China alone took 352 billion. Europe drew 137. North America and Oceania, 122. Then the numbers fall away — 21 billion for the Middle East and North Africa, 18 for Eurasia, 18 for all of Sub-Saharan Africa. Capital is flowing but mostly to a handful of places. Next, let's bring the view back to the United States. Figure three. What the United States actually runs on, according to the International Energy Agency. Total final consumption for 2023: oil products at 47.2 percent, natural gas at 23.9, electricity at 21.4, biofuels and waste at 6.1. Coal is down to eight tenths of one percent. The second panel traces those same lines back to the year 2000, and the shape is the point. Twenty-three years, and the mix has moved far less than the conversation around it. This brings us to the chart that defined the year: artificial intelligence demanded new power. The next two figures come from S&P Global. Figure four. Data centers arrive on the grid. S&P's forecast, stated in its own headline: US power demand from data centers is expected to more than double from current levels. The bars sit low and flat from 2018 through 2022. Then the shaded forecast zone begins and the climb turns steep, straight through 2030. This is the chart behind the sentence "the scale of compute demanded new power." A demand curve does not wait for the interconnection queue to be ready. Figure five. The same demand, placed on a map. S&P plots every US data center campus drawing at least one megawatt of utility power — blue for operating, orange for planned, each circle sized by megawatts up to 1,500. Two things jump out. The orange sits on top of the blue, which means new load is concentrating where load already is, rather than spreading. And both magnified insets pull in on PJM — the grid operator covering the mid-Atlantic — where the crowding is heaviest. Here, we see concentration risk. The demand requires a vast physical build. McKinsey tries to put a price on it. Figure six. McKinsey prices the build. Energy, power, and resources: 23 trillion dollars of estimated investment by 2040. Developing countries account for 85 percent of new demand. Renewable projects are scaling fast enough to supply the majority of power generation. And stakeholders are fortifying vulnerable grids and investing heavily in new transmission. The generation question is largely answered. The delivery question is still open. Figure seven. The other half of the same bill. Digital infrastructure: 19 trillion dollars by 2040. The third takeaway is the one worth hearing twice — power access is constraining data center construction, and that constraint is driving a wave of joint ventures at the intersection of energy and digital infrastructure. Two industries that used to buy separately now sit at the same table. Put figure six and figure seven side by side and the estimate runs to 42 trillion dollars by 2040. The eighth chart may be my favorite. I think of it as the system-thinking machine. It comes from DNV, the assurance and risk-management firm. DNV's Energy Transition Outlook model, drawn as a single diagram. Transport, power supply, buildings, hydrogen supply, manufacturing, fossil fuel supply, and carbon capture all wire into energy demand at the center. Blue arrows carry prices. Green arrows carry policy. It looks like a tangle because it is one — nearly every box feeds a box that feeds it back. This is what a system thinking machine looks like when someone draws it honestly. The final two charts, from J.P. Morgan, let's take a look. Figure nine. Six waves of innovation, and an open question. Their timeline runs from the 1800s. First wave: Iron, waterpower, and mechanization. Second wave: steam engines, steam power, and rail. Third: electricity, steel, and heavy engineering. The 4th wave involves mass-produced automobiles and cheap oil. Entered the 5th wave: information technology, telecommunication, and software. Today, we are in the 6th wave that is marked as industrial strategy, infrastructure, and supply networks. Britain led the first two waves. The United States led the next three. The current wave carries a big question mark where the flag should be. Figure ten. Who holds the inputs. The reserve map underneath, drawn from the U.S. Geological Survey. China holds 90.9 percent of global gallium and 20 percent of zinc. South Africa, 77.8 percent of platinum group metals. The Democratic Republic of the Congo, 54.5 percent of cobalt. Chile, 31 percent of lithium; Australia, 23.3 percent. Guinea, a quarter of the world's bauxite. Every build in every other chart here depends on this one. Which brings back the question that runs under all ten charts: who carries the risk when a single country holds ninety percent of an input you cannot substitute? Taken together, the ten charts tell one story. Emissions remain above their pre-pandemic level. Capital is flowing, although unevenly. Artificial intelligence is pulling electricity demand upward faster than grids and supply chains can adjust. Meeting that demand will require trillions of dollars across energy and digital infrastructure, and every part of the build depends on a concentrated set of minerals, markets, and countries. That brings me back to the three thoughts at the beginning. Systems rebalance. Speed needs judgment. And the people who can see across energy, climate, artificial intelligence, and infrastructure will be best equipped to lead through what comes next. Progress has a direction. Stewardship determines where it lands. That's the piece. The written version, with all ten charts includes links to every source, is at cleanpowerwhisperer.ai. If any of this is live in your own portfolio and you'd like to think it through with someone, you know where to find me. . Stay curious. Be safe. Be well.