The most important infrastructure revolution of our time is happening almost invisibly. While politicians gather in Brazil for COP30 to debate climate targets and activists still occasionally glue themselves to artworks, market forces combined with artificial intelligence are fundamentally reshaping how we generate, distribute, and consume energy. And here’s the twist: it’s working precisely because it’s driven by economics, not evangelism.

At NEXT24, when Arash Aazami introduced the concept of “Energy Buddhism” and we explored AI as plumbing rather than product, we were identifying threads that now weave together into a compelling narrative. Three recent analyses – from Bill Gates, Michael Liebreich at BloombergNEF, and The Economist on China’s clean energy dominance – together with NEXT24 insights reveal something unexpected: the energy transition is succeeding not through sacrifice, but through superior economics enabled by invisible AI infrastructure. The IEA’s latest analysis confirms the trajectory: renewables will meet over 90% of global electricity demand growth through 2030 and surpass coal as the largest generation source by the end of 2025.

When plumbing becomes revolutionary

Remember our observation from NEXT24 that AI is best understood as plumbing? Not the glamorous, product-facing ChatGPT variety, but the invisible infrastructure that enables previously impossible things. Indoor plumbing transformed public health. The internet transformed commerce and culture. AI is now transforming our ability to manage complexity at scales humans simply cannot comprehend.

Consider: Germany now has over a million balcony solar installations. Add rooftop systems, community solar, wind turbines, and you’re looking at millions of distributed energy generators. This isn’t your grandfather’s electrical grid with a dozen massive power plants feeding passive consumers. It’s an internet of energy with prosumers – people who both produce and consume – creating a mesh network of unprecedented complexity.

Managing this? Humanly impossible. For AI? Trivially easy.

When everyone in a neighbourhood comes home at 6 PM, plugs in their electric car, turns on the cooker, and the heat pump kicks in, something has to give. A century ago, grid operators would have built bigger power plants. Today, AI can orchestrate millions of micro-decisions: pause the heat pump here, delay EV charging there, draw from home batteries elsewhere, pre-heat buildings when solar is abundant. The intelligence layer makes the system work.

This is Physical AI – the application of machine learning techniques not to text and images, but to sensor data from the physical world. It’s what Ivan Poupyrev identified at NEXT24 as the next frontier: understanding and optimising reality itself.

The economics of energy that everyone missed

Here’s where the market forces become irresistible. Michael Liebreich’s analysis at BloombergNEF cuts through the noise: we’re at or near peak emissions not because of heroic sacrifice, but because clean energy has become cheaper. One in five cars sold globally is electric. Clean energy supplies over 30% of global energy services. Wind and solar costs have plummeted below those of fossil fuels in most markets. The IEA calculates that deployment of solar PV, wind, nuclear, electric cars and heat pumps since 2019 now prevents 2.6 billion tonnes of CO2 annually – equivalent to 7% of global emissions.

The math is simple: if clean energy grows just 3% faster than total energy demand – and it’s currently growing far faster – fossil fuels get squeezed out inexorably. No drama required. Just patient, compounding mathematics.

Michael Liebreich captures the inevitability with a historical analogy:

In the 11th century, after his failed attempt to stop the waters, King Canute declared that “all the inhabitants of the world should know that the power of kings is vain and trivial.” President Trump and Energy Secretary Chris Wright should take note.

But here’s the part that Bill Gates emphasises – and that the climate community often resists: we should measure progress by human welfare, not temperature targets. In his recent essay to COP30 attendees, Gates explicitly rejects climate “doomerism”:

Although climate change will have serious consequences—particularly for people in the poorest countries—it will not lead to humanity’s demise. People will be able to live and thrive in most places on Earth for the foreseeable future.

Energy demand will double by 2050, and from a human flourishing perspective, that’s positive. More energy means economic growth, which means prosperity, which is the single best defence against climate impacts.

The controversial bit? Gates argues that developing nations should use natural gas if it accelerates their development. The warming effect is vastly outweighed by the benefits of lifting populations out of poverty. Economic development isn’t opposed to climate solutions – it’s essential for them, he argues.

China’s accidental green energy gift to the world

The Economist‘s analysis of China’s clean energy dominance reveals something unexpected: by achieving manufacturing scale that dwarfs all competitors, China has made clean energy abundant and affordable globally. Not through altruism, but through industrial policy married to market economics.

China’s numbers are almost incomprehensible:

• 887 gigawatts of solar capacity (nearly double Europe and America combined)
• Manufacturing capacity projected to exceed global climate targets by 65%
• Costs falling so fast that emerging markets are leapfrogging developed nations in electrification

Nepal went from essentially zero EVs to 76% of new vehicle sales in just a few years, thanks to cheap Chinese imports. African villages are replacing diesel generators with solar, not because of aid programmes, but because solar is now simply better economics. A quarter of developing economies now exceed American levels in electrification of end-use sectors.

This is Payal Arora’s insight from NEXT24 playing out: the Global South is charging into these technologies with more optimism than the West because they have less legacy infrastructure to protect. They see the upside more clearly because the downside weighs less heavily.

The AI energy paradox

Now comes the delicious irony. AI – specifically training large models and running inference – is an energy glutton. Data centres are projected to double their electricity consumption by 2030, adding demand equivalent to several entire countries.

Climate activists immediately panicked: Won’t this derail the transition?

The opposite is happening.

AI companies like Microsoft, Google, and Meta have net-zero commitments and profit margins that can absorb green premiums other industries cannot. They’re signing massive power purchase agreements for renewable projects, driving investment in grid-scale batteries, even reviving nuclear power (though Germany has notably taken a different path, phasing out nuclear to pursue renewables-only decarbonisation). Microsoft is restarting Three Mile Island. Google is partnering with advanced nuclear startups.

But the real magic is how AI’s demand for reliable, 24/7 power is forcing exactly the infrastructure investments that make distributed renewable grids work:

• Grid-scale battery deployment
• Smart grid management systems
• Demand response technologies
• Advanced weather prediction for renewable output
• Optimisation algorithms for millions of prosumers

Every improvement in fundamental infrastructure driven by AI’s energy appetite benefits all energy users. The learning curve accelerates. The costs fall faster. The infrastructure gets built that unlocks broader electrification.

It’s a virtuous cycle masquerading as a vice – what David Mattin at NEXT23 called “the most important technology symbiosis ever.” Or, as Ben Thompson argues in his analysis of the benefits of bubbles, what looks like irrational AI speculation is actually a coordinating mechanism driving infrastructure investments that will outlast the hype.

The three energy mindsets we actually need

Arash Aazami’s “Energy Buddhism” framework captures something the purely economic analyses miss: transformation requires new ways of thinking, not just new technologies.

• The Seeker’s Mindset recognises that in nature, nothing is lost, only transformed. Growth is transformation. Those African villages switching from diesel to solar aren’t sacrificing – they’re upgrading. The scientific mindset sees opportunity in change.

• The Artist’s Mindset brings creativity to constraints. The Sunglacier project used solar panels to create ice in the Sahara Desert – funded as a conflict-reduction measure because water abundance eliminates the need to fight over it. Artists reveal possibilities engineers overlook.

• The Hacker’s Mindset uses existing reality to create desired reality. The internet went from 63 card-carrying nerds in 1973 to five billion users in 2022 by making complexity disappear behind simple interfaces. The energy transition needs the same approach: make the complex simple, the expensive cheap, the difficult easy.

These aren’t just inspirational platitudes. They’re practical frameworks for implementation. The seeker sees that fossil fuels are the historical anomaly, not renewables. The artist imagines plant-powered lights and solar-film clothing. The hacker builds unify.energy to create self-balancing energy meshes.

What this means for business

The implications ripple outward:

• For technology companies: Your energy strategy is no longer optional or peripheral. You’re driving the infrastructure build-out whether you realise it or not. The question is whether you’re strategic about it. Long-term power purchase agreements (PPAs) lock in costs. Colocation with renewables reduces expenses. Participation in demand response monetises flexibility.

• For manufacturing: Electrification is coming for your processes. The companies figuring out heat pumps for industrial heat, electric arc furnaces, and process electrification are building competitive moats. Chinese manufacturers have understood this for a decade. Western firms are finally catching up.

• For automotive: The BEV transition has reached escape velocity. Based on 2020-2024 CAGR trends from ACEA data, Europe is on track for 65-70% BEV market share by 2035 despite recent policy headwinds. The question isn’t whether to electrify – it’s how to capture value in the transition. Fleet services, charging networks, and grid integration offer larger margins than vehicle sales.

• For energy companies: The flexible gas opportunity is real but time-limited. As Liebreich argues, getting to 90% renewable grids is economically straightforward. The last 10% needs flexible backup, creating a role for gas that can provide grid services even when not generating power. The window is perhaps 15 years before batteries or long-duration storage close that gap.

• For everyone: The prosumer economy is emerging faster than most realise. Rooftop solar, home batteries, EVs with vehicle-to-grid capability – these aren’t futuristic concepts. They’re available today in many markets. The regulatory barriers are falling. The economics are increasingly compelling.

The pragmatic path forward on clean energy

What separates useful analysis from wishful thinking? Michael Liebreich’s framework offers clarity:

Focus on the 90% of decarbonisation that’s economically viable today. Stop obsessing over perfection. Accept that flexible gas will play a role in grid stability for years to come. Reject hydrogen fantasies – green hydrogen has no plausible pathway to affordability for most applications. Shift levies from electricity to fossil fuels so heat pumps can compete on operating costs. Embrace locational pricing to drive efficient investment.

Most importantly: stop treating citizens as “hostages at a cash machine”, forcing them to fund priorities they don’t share through regressive energy taxes. Win the sensible centre by making clean energy cheaper and better, not by making fossil fuels more expensive and worse.

And here’s the secret that makes this strategy work: technologies like heat pumps and BEVs aren’t just winning on economics and emissions – they’re often superior products. Electric vehicles deliver instant torque, near-silent operation, and minimal maintenance. Heat pumps provide both heating and cooling from a single system with dramatically lower running costs. When the better product is also the more economical and ecological choice, you don’t need mandates or guilt. You just need to remove barriers and let consumers choose.

This isn’t climate denialism or delayism. It’s pragmatism. The transition is happening faster than almost anyone predicted, driven by cost curves and market forces. The role of policy is to remove barriers and align incentives, not to mandate outcomes.

The invisible infrastructure thesis

We return to where we started: the most important infrastructure revolutions are invisible.

You didn’t wake up one morning thinking “I need TCP/IP in my life.” But that invisible protocol layer of the internet enabled everything that followed – e-commerce, social media, streaming entertainment, remote work, the app economy.

AI is becoming that invisible layer for the physical world. It’s the protocol that enables millions of distributed energy sources to act as a single coherent system. It’s the intelligence that turns buildings into thermal storage devices, cars into mobile batteries, and neighbourhoods into self-balancing microgrids.

The energy transition isn’t failing. It’s succeeding in precisely the way our analysis at NEXT24 predicted: through invisible infrastructure enabling market-driven transformation, powered by superior economics, and accelerated by AI’s voracious appetite for the very infrastructure that makes broader electrification possible.

The fossil fuel age, which Aazami correctly identified as a historical blip spanning just two centuries, is ending not with a whimper or a bang, but with a spreadsheet. Because the numbers have shifted. Clean energy is cheaper. AI makes complexity manageable. Chinese manufacturing makes supply abundant. And human welfare – not temperature targets – provides the moral compass.

What comes next for the energy transition

As delegates gather in Brazil for COP30 to negotiate emissions targets and climate finance commitments, the real transformation is happening elsewhere – in Chinese solar factories, European EV showrooms, AI data centres signing renewable PPAs, and millions of rooftop installations. The next decade will be defined by implementation, not negotiation. We have the technologies. The costs are falling fast enough. The Chinese manufacturing base can supply the world. What remains is deployment at scale – what Ezra Klein and Derek Thompson call the shift to an “abundance agenda” in their 2025 book Abundance: ending artificial scarcity by actually building infrastructure.

This requires:

• Removing regulatory barriers to prosumer participation
• Building grid intelligence into everything
• Accepting that perfect is the enemy of good
• Measuring success by human flourishing
• Letting markets drive adoption through economics
• Using policy to align incentives, not mandate outcomes

At NEXT24, we talked about the need for new energy everywhere – in our work, our creativity, and our systems. What we were really describing was a fundamental shift in how we think about and implement transformation.

The climate conversation has been dominated by guilt, fear, and sacrifice. But transformation driven by those emotions rarely succeeds. What works is transformation driven by self-interest, opportunity, and better outcomes.

The invisible infrastructure revolution is winning because it offers superior economics wrapped in smarter technology enabled by market forces. That’s not a compromise with our values. That’s how durable change actually happens.

The plumbing revolution continues. And just as with indoor plumbing a century ago, future generations will look back, amazed, that we ever did things the old way.

Picture by Aline Massuca/COP30