Mining sits at the centre of the defining economic tension of the decade, and very few in the investment world are paying it proper attention. The world needs dramatically more metals to electrify, digitalise, and rebuild its infrastructure. The mining industry, as it stands, cannot deliver. This piece looks at that the existing gap, how we got here, and why a new generation of AI-powered optimisation software may be the most immediate solution available.

---

Commodity markets just had one of their most violent weeks in decades.

Gold touched $5,580 an ounce on January 30th, its highest level in history, before suffering its steepest one-day drop since 1983 and falling 21% by Monday. Silver lost 41%. Copper shed 11% from record highs. The trigger was Trump’s nomination of Kevin Warsh as the next Fed Chair, which strengthened the dollar and forced a violent unwind of leveraged positions. By mid-week, gold was clawing back toward $5,000, with JPMorgan reaffirming a $6,300 year-end target. The structural bull case, they argue, remains intact.

But the whiplash in precious metals obscures a more important story. While gold and silver were crashing on macro sentiment, the industrial metals that actually power modern infrastructure have been on a different trajectory. Lithium surged nearly 30% in January as Chinese supply restrictions tightened the market. Cobalt has more than doubled since the DRC imposed export quotas that halved shipments. Copper, even after last week’s correction, remains near record levels with structural deficits projected to widen through the decade.

The precious metals sell-off will dominate headlines. But the tension between what the world needs from mining and what the industry can deliver is the story worth paying attention to.

Why mining matters more than you think

An electric vehicle requires roughly six times more critical minerals than a conventional car. A single offshore wind turbine contains around 15 tonnes of critical minrals. The electricity grids that need to double in capacity run on copper, aluminum, and rare earth elements.

The IEA projects that mineral demand for clean energy needs to triple by 2030 and quadruple by 2040. Copper faces a projected supply deficit of 150,000 to 330,000 tonnes in 2026 alone, and potentially millions by decade-end. Goldman Sachs sees copper reaching $15,000 per tonne by 2035. J.P. Morgan projects $12,500 by mid-2026. And this is before accounting for AI: data centres are copper-intensive, and J.P. Morgan estimates AI-related copper demand could reach 475,000 tonnes in 2026, roughly equivalent to adding another medium-sized mining nation to the demand side.

The entire energy transition, and increasingly the digital infrastructure build-out, rests on whether we can dig enough material out of the ground, fast enough, at a cost the world can afford.

But mining is broken

Mining is one of the only major industries on Earth where productivity has gone backwards. McKinsey put a number on it: productivity has halved since 1997. Not stalled. Halved. Manufacturing more than doubled over the same period. Agriculture improved by 50%. Mining went the other direction.

The reasons are structural. Ore grades are declining: global copper grades have fallen from 1-2% historically to below 0.7%, forcing miners to process exponentially more rock per tonne of metal. Mines are deeper, haul distances longer, processing more energy-intensive. On top of that, the industry spent a decade underinvesting after the commodity supercycle peaked around 2011. Fewer than ten significant copper discoveries have been made in the past decade. Refined copper output growth is expected to slow to just 0.9% in 2026, down from 3.4% in 2025.

The result: a sector asked to deliver unprecedented output with deteriorating geology, ageing assets, and insufficient capital.

How we got here

Mining has always been geographically concentrated, dictated first by geology and then by politics. For much of the twentieth century, Western companies dominated. That changed as resource-rich nations asserted sovereignty and economics shifted toward lower-cost regions. Decades of underinvestment in Western mining capacity, driven by environmental opposition and the comfortable assumption that cheap imports would always be available, reinforced the pattern.

China filled the gap, not primarily through mining but through refining and processing. Today, China controls over 60% of global lithium and cobalt processing, over 90% of graphite and rare earth refining, and leads on 19 of 20 strategic minerals tracked by the IEA. When China began imposing export restrictions on critical minerals in late 2023, starting with gallium and graphite and escalating through 2025 with rare earths and tungsten, the vulnerability became impossible to ignore. Trump invoked the Defence Production Act in March 2025 to fast-track domestic production. The EU mandated domestic processing targets. The US created a $2.5 billion strategic minerals stockpile.

But even if permitting were instantaneous, opening a new mine takes 15 to 20 years. The supply chain cannot be reshored overnight.

The case for optimising what you have

If new supply is a decade away and demand is accelerating now, the most immediate lever is to extract more from what already exists.

A mid-size mine generates terabytes of data daily from sensors on grinding mills, flotation circuits, conveyor belts, and processing plants. Historically, less than 1% of that data has been meaningfully analysed. A new generation of software platforms is changing this, building AI-driven systems that optimise operations in real time: adjusting grinding parameters to match ore variability, controlling flotation chemistry, managing stockpile blending, and scheduling dispatch dynamically. Deployments across copper, gold, and platinum operations have delivered throughput improvements of 3-5%, energy savings of 5-8%, and recovery gains worth tens of millions of dollars annually per site.

What makes this vertical most interesting is the moat. These platforms are built on physics-informed machine learning, combining metallurgical constraints, geological variability, and process engineering with statistical learning. The training data required is a decade of site-specific operational history. Integration involves connecting 15-20 OT and IT systems. Switching costs, once a platform is embedded in a mine’s control loop, are prohibitive.

This is vertical AI at its most defensible. And the companies that built their technical foundations before the current AI wave, accumulating years of domain-specific data and deployment experience, hold a structural advantage over later entrants trying to build from scratch.

Where this leaves us

The commodity volatility will continue. Last week proved that. But the sell-off in precious metals was a positioning unwind, not a fundamental shift. The recovery in battery metals is driven by structural demand (electrification, grid expansion, energy storage) colliding with supply constraints that are finally biting.

Mining faces pressures that are not cyclical but structural: declining ore grades, rising costs, geopolitical fragmentation, and an unprecedented demand shock from the energy transition and digital infrastructure. The companies that navigate this will be the ones that figured out how to do more with less.

Technology that makes mining more efficient is not an optional upgrade. It is becoming a prerequisite. And for investors looking at the intersection of industrial digitisation and the energy transition, it is one of the more compelling places to be paying attention.

1. Reuters via Kitco News, ”J.P. Morgan expects gold prices to reach $6,300/oz by end of 2026,” February 2, 2026. https://www.kitco.com/news/off-the-wire/2026-02-02/jp-morgan-expects-gold-prices-reach-6300oz-end-2026
2. International Energy Agency, Global Critical Minerals Outlook 2024, Executive Summary. https://www.iea.org/reports/global-critical-minerals-outlook-2024/executive-summary
3. McKinsey & Company, ”Unearthing a new era of innovation in mining,” December 9, 2025. https://www.mckinsey.com/industries/metals-and-mining/our-insights/unearthing-a-new-era-of-innovation-in-mining