I love the discipline that writing forces on me. In a world moving at breakneck speed, if you can’t think clearly, you can’t write clearly. That’s part of what makes generative AI so powerful—you can now use language itself as an input to create machine-based output. No longer do you need to be able to code to make things happen in the digital world.
But while this letter is entirely my own writing (minus some fact and spell-checking AI help), I’ll be honest: there’s no way to predict the future. What we can do is build a research process that gets us as close to the truth as possible—understanding what’s really happening in AI and energy, both on the ground and at scale.
At Nadia Partners, we originate companies from zero to one to scale. We focus on AI and energy because these two forces are deeply intertwined, reshaping both the digital and physical world. We don’t just invest in companies—we start them. We find constraints, build businesses around them, and scale from there.
I’ve always admired Warren Buffett—not just for his investing genius but for how he simplified his business life: own great companies for a long time. But what if, in the age of AI, we could build a collective of companies working together to solve the biggest problems in AI and energy? That’s exactly what we’re doing.
This letter isn’t about what’s trending. It’s about what we believe—where we have conviction. And conviction, in our world, comes down to one word: constraint.
The past 24 months have been unlike anything we’ve seen before. AI is moving at a speed that outpaces every major technological shift of the past—whether it’s the internet, cloud computing, or mobile. But this time, there’s one key difference: a person in Ethiopia now has access to intelligence at the same speed as someone in New York. That has never happened before. AI’s reach is immediate, its impact universal. The world is still underestimating AI’s effect on business, geopolitics, and energy systems. There’s a massive gap between those talking about AI and those actually building with it. AI is not some abstract concept anymore—it’s shaping capital flows, national strategies, and corporate structures at an astonishing pace.
One of the biggest misconceptions is that AI is just a digital phenomenon—that the impact is limited to software, enterprise automation, and large language models. That’s completely wrong. AI is not just consuming energy; it’s reshaping how energy is produced, stored, and delivered. Right now, AI models are hungry. GPUs are devouring electricity at unprecedented rates. The growth of AI-native infrastructure—data centers, high-performance computing clusters, and on-premise inference—has already put enormous strain on the grid. But the real wave of AI-driven energy consumption is just getting started.
The next major AI consumers won’t be just data centers. They’ll be humanoid robotics, AI-driven factories, and self-learning industrial systems. For decades, energy demand was predictable. Not anymore. AI is pushing power infrastructure to its limits, and we don’t have enough supply to meet future demand. Data centers, already consuming around 2% of U.S. electricity, are projected to rise past 5% soon. Electric vehicles are placing massive new loads on the grid. Industrial electrification—factories, automation, AI-powered logistics—is scaling rapidly. And at the same time, our grid is outdated, built for a world that didn’t anticipate the scale of AI’s energy needs.
There’s an unavoidable truth that few are talking about: power is becoming more constrained, more valuable, and harder to come by. This is why origination matters. In energy, if you don’t own the right projects in the right locations, you won’t have access to power. That’s why we focus on unlocking new energy projects before the market realizes their potential, building AI-optimized data centers, and expanding transmission and storage to handle AI-driven loads. But solving for power isn’t enough. Because energy is just one part of the equation. AI is about to force a complete rethinking of how physical industries operate.
While energy is the constraint, AI is the unlock. We’re entering an era where companies won’t just use AI; they will be built with AI as a native function. These businesses will be fundamentally different from traditional software or industrial firms. One of the most overlooked aspects of AI’s rise is its ability to optimize its own energy consumption. AI isn’t just consuming power—it’s actively reshaping energy management in real-time. AI-powered grid optimization is unlocking massive efficiency gains, reducing waste, and balancing power loads in ways humans never could. AI-driven industrial automation is shifting power demand away from peak hours, smoothing out energy usage across the day. Autonomous factories are creating entirely new load profiles for energy systems, operating on self-optimized schedules that minimize costs and maximize efficiency.
This is why the next industrial revolution won’t just be about software. It will be about the AI-native industrial stack—where AI drives both software and real-world automation. AI-native businesses will challenge fundamental economic assumptions. Labor costs, capital allocation, and supply chain constraints will all be rewritten. The industries that have traditionally relied on human-intensive operations—manufacturing, construction, logistics, and energy production—will become AI-first industries. This changes everything.
The way businesses are built is about to change. The old model—starting one company at a time, scaling slowly, and treating each company as an isolated entity—is too slow. The next generation of companies will be loosely connected via APIs, launching in cohorts across different industries, each forming its own collective. Instead of a single company trying to solve an AI-energy problem, a network of companies will launch in parallel, each focused on a different piece of the puzzle. These businesses will interact dynamically, much like how microservices communicate in software development.
Imagine a world where AI-native energy trading firms optimize power distribution in real-time, using predictive AI models to balance energy supply and demand. Autonomous construction startups build AI-powered factories and infrastructure at record speeds. AI-driven logistics companies reshape global supply chains with predictive automation and dynamic routing. Each of these businesses is independent, but they function as a network, compounding value together. This is the future of company-building: cohorts of AI-first businesses launching in sync, solving interconnected problems at scale.
At Nadia Partners, our focus is simple: originate, build, and scale. We don’t chase trends—we identify constraints and build companies to unlock them. AI is accelerating at an insane pace. Energy is becoming the defining bottleneck of the 21st century. The companies that recognize this—and build AI-native solutions in both digital and industrial sectors—will define the next decade. The winners in this new era will not be those who simply apply AI to existing business models, but those who rethink entire industries from the ground up.
We’re all in. The unlocks are only beginning.
Aidan Kehoe
Founder and Managing Partner
Nadia Partners