OPINION: The AI future will require unprecedented amounts of power, and embracing nuclear energy is an obvious, clean solution that could boost the NZ economy for generations. But do we have a leader with the courage? By Thomas Scrimgeour
Artificial Intelligence is transforming our world, though not in the way most people imagine.
While the knowledge industry revolution is still around the corner, the warehouse-sized computers driving this innovation can’t be built fast enough.
Data centres already consume roughly 2% of global electricity, which is more than 10 times New Zealand’s annual generation, and this figure is projected to double by 2026.
Elon Musk’s xAI recently built the world’s largest supercomputer, “Colossus,” in 122 days. They then doubled its size in just 92. It now requires the power of a small city to operate, and xAI is turning to non-renewables to supply this insatiable need.
Around the world, AI’s energy demands are rising faster than clean energy capacity can keep up. In Northern Virginia, a major data centre hub, AI-driven power use is expected to triple by 2029, while clean energy capacity will only double.
An opportunity for New Zealand
This problem is our opportunity. Countries able to deliver clean, reliable, and affordable energy will be best placed to attract billions in data centre investment.
The good news is that we already have a distinct competitive advantage. Nearly 90% of New Zealand’s electricity generation is renewable, our temperate climate lowers cooling costs, and we’re politically stable with strong privacy protections. The sales pitch writes itself.
Global hyperscalers have already noticed. Microsoft has invested $1 billion in New Zealand data centres, and Amazon Web Services plans to spend $7.5 billion on their new data centre in northwest Auckland.
But here’s the hitch: we might have what the world wants, but we don’t have enough of it.
Hydroelectricity is great, but we’re not about to dam another river. Wind and solar are neat, but in midwinter they contribute very little. When renewables fall short, coal and gas fire up, bringing last winter’s power price headlines back to haunt us.
We need more generation and innovation. The big four power companies, known as gentailers, both produce and sell electricity to consumers. They aren’t investing enough in new generation and critics argue the market incentives aren’t there to expand capacity. Paul Fuge from Consumer NZ site Powerswitch puts it bluntly: “the results we’re seeing aren’t what you’d expect from a thriving competitive market.”
Market reform could help. But the real opportunity lies in increasing power production.
The morning’s headlines in 90 seconds, including Zelensky’s suit becomes hot topic at peace summit, a cold blast on the way, and Auckland FC lures new signing back home. (Source: 1News)
Conventional geothermal is our best near-term lever. It already supplies nearly 20% of our electricity and operates 24/7, unlike weather-dependent renewable energy. The best estimates suggest that we have enough active geothermal zones to double our output.
I believe supercritical geothermal is the natural next step. It involves drilling five kilometres into the Earth’s crust to unlock ten times the power of conventional geothermal. However, supercritical geothermal still faces significant technological hurdles. Commercialisation isn’t expected until the late 2030s, and it’s unclear how quickly it could scale.
So we need a second pillar of clean energy generation, which brings us to the last swear word in New Zealand politics.
Nuclear. (I can already smell the uranium.)
Can we turn around the taboo?
It’s only a strange quirk of history that nuclear power is controversial in New Zealand. It got bundled together with the protest backlash of the 1980s, and we’ve never quite moved on. The 1978 New Zealand Royal Commission on Nuclear Power was expecting a “significant nuclear power program in the early part of next century.” Better late than never, I suppose.
Although traditional nuclear power is brilliant, high upfront costs and a long build time put it in the too-hard basket, especially given New Zealand’s basic revulsion.
But nuclear technology is rapidly evolving. Small Modular Reactors (SMRs), one-tenth the size of conventional plants, are on the horizon and could be installed in a fraction of the time. A bold government could break the nuclear taboo in a single term.
The Americans want SMRs by the end of the decade. One company, NuScale, already has regulatory approval. Canada will build four 300-megawatt reactors by the mid-2030s, and Japan is reversing plans to decommission its nuclear power plants.
All we need is a leader with the courage to take the first step.
A feasibility study to work out the who, when, and where of SMRs could be started today. The first politician to raise the issue will take some heat, but Kiwi voters will reward conviction and enjoy the benefits for generations.
Energy abundance is the foundation of every productive economy, and the only road to lasting prosperity.
We have the chance to do two big, good things: create a data centre industry for New Zealand and generate enough power to bring down costs for everyday Kiwis. But opportunities like this don’t wait around.
The time to act is now.
Thomas Scrimgeour is a reasearcher at the Maxim Institute, an independent think tank based in Auckland.
