Tech-Rocks Pt.3: How Commodities And AI Lead to More of Both

Here is my call for 2026 based on yesterday’s four scenarios.

Yesterday, I presented 4 scenarios for markets in 2026:

Scenario 1: Commodities win, AI stays “dead”.

Scenario 2: AI comes roaring back.

Scenario 3: Both rip higher.

Scenario 4: Everything melts down.

It’s scenario 3 for me.

With the potential for scenario 4 towards the tail end of the second half of the year.

(I’ll discuss drivers of that potential event next week AND what history shows us about the best strategies to navigate it)

But for now, let’s look how AI and commodities can both rip higher at the same time.

Not one or the other. Both.

The market will oscillate between themes that dominate week to week. But I think the underlying trajectory for both is up.

It’s a powerful market setup. The demand is too real. The supply is too constrained.

Here’s how everything fits together.

AI cannot scale without commodities

A single large hyperscale AI data centre uses 50,000 tonnes of copper.

Cooling systems account for about 40 percent of copper usage in AI-focused data centres.

Hundreds of facilities are planned globally.

Lithium-ion batteries back up every one of these installations.

AI systems cannot tolerate power disruptions. Even a millisecond outage corrupts training processes and wastes millions in compute.​

Rare earth neodymium magnets power the cooling fans in GPU-intensive server racks. Each cooling fan contains ~50 to 150 grams of neodymium-iron-boron material.

Graphite anodes make up 95 percent of lithium-ion battery anode material.

China produced 92 percent of global spherical graphite and 99 percent of synthetic graphite in 2024.

The buildout is locking in decades of commodity demand.

Commodities cannot scale without AI, and AI is helping find commodities

The irony is that there’s a virtuous feedback loop that cuts both directions.

The same infrastructure creating commodity hunger is now being deployed to find those exact materials faster.

KoBold Metals uses AI and machine learning to hunt for lithium and cobalt.

Fleet Space Technologies combines real-time 3D seismic imaging with AI-driven target recommendations.

To name just a few of the data-centric and AI enabling explorations approaches I’ve looked at in the market.

Not least of which is current Australian Small Cap Investigator stock and mining data services and sensor provider, Imdex [ASX:IMD]:

Source: Market Index

[Click to open in a new window]

***Note: IMD is well past its price at recommendation, this is not financial advice, simply an example***

What this means for 2026

AI cannot scale without massive commodity inputs, which means every data centre expansion directly tightens supply for copper, lithium, and rare earths.

Commodity discovery is accelerating because of AI, with machine learning slashing exploration costs and improving success rates.

Supply chain verification runs on the infrastructure that needs verification, creating a closed loop where AI-powered traceability systems authenticate the minerals feeding AI infrastructure.

The trick is finding the companies sitting at this intersection: Australian miners with clean jurisdiction and tech-enabled discovery capabilities that can command premium pricing for traceable supply.

Battery developers using AI-optimized chemistries to reduce material intensity and improve performance metrics.

Better chemistries improve production economics. Increasing battery market uptake and accelerating the rollout. Which, in turn, feeds the demand for material inputs.

So that’s the call: both themes eventually rip in 2026 because they are locked together.

The market will rotate between them, but the underlying trend is up for both.

Let’s see how the call plays out.

Next week, when you hear from me, I’ve got more on the currency angle behind the commodity-AI nexus and its role in potential financial doomsday scenarios for 2026.

And what you can do about it.

There’s plenty under the hood.

Talk then.