Robert Friedland, founder of Ivanhoe Mines and one of the most consequential mining entrepreneurs of the last four decades, sits down with Bloomberg's Odd Lots hosts Joe Weisenthal and Tracy Alloway at Bloomberg's New York headquarters to deliver what amounts to a sustained warning about one of the most consequential supply crises in modern economic history.1 The conversation, recorded September 10, 2026, with copper trading near record highs of $14,875 per ton, covers the structural failure of global copper supply, the voracious new demands of artificial intelligence and electrification, the weaponization of critical raw materials, the quiet deindustrialization of the United States, and the near-impossibility of solving any of it with yesterday's technology. The hosts frame the episode with appropriate context: global production of copper is on track for its first annual decline since 2017, even as prices surge. That paradox is exactly where Friedland begins.
10,000 Years in 18
The opening thesis lands without preamble. "Standard & Poor's has declared copper as the new oil," Friedland says. What follows is a quantification of the challenge that should stop investors cold: "We need to mine as much copper in the next 18 years as we did in the last 10,000 years just to maintain global 3% GDP growth without the electrification of the world economy."
That number is not rhetorical. It is the arithmetic consequence of population growth, global economic expansion, the physical demands of electrification, and the accelerating appetite of data infrastructure. Friedland makes clear that when you add electric vehicles, grid modernization, and AI-driven data centers to the equation, the demand curve does not flatten. It steepens.
The supply side of that equation, however, is moving in the wrong direction.
The Mine That Is Dying
To illustrate the structural deterioration of copper supply, Friedland turns to La Escondida in Chile, the world's largest copper mine, owned by Rio Tinto and BHP. "When it was first found, a generation ago, the grade was close to 2% copper," he explains. "Today it's 0.8 of 1%." Within two years, that grade falls to 0.4 of 1%. As the ore degrades, the energy required to extract the same amount of metal climbs exponentially.
The physics are unforgiving. "You need to take rock and reduce it to talcum powder to get the metal out," Friedland says. "About 4 to 5% of all the energy produced on planet Earth is utilized in the crushing and grinding of rock." La Escondida sits at 12,000 feet in the Atacama, a desert that has not seen rain in 30,000 years. A $4 billion desalination plant now pumps ocean water uphill at a cost of a million dollars a day in electricity. The mine is spending 10 to 12 billion dollars to produce less copper than it did a decade ago. "Our global production of copper this year is down even though prices are up," Friedland observes. "And that is what's happened to our industry."
AI's Insatiable Appetite
The demand side accelerant is artificial intelligence. Friedland offers a comparison that reframes the energy transition debate entirely. A standard Google search consumes the electrical equivalent of running a 100-watt bulb for twelve seconds. "When you do an AI search," he says, "you're using thirty times more electrical energy." Every child on the planet with an iPad will run AI searches within years. "The electrical demand for AI literally goes to infinity," Friedland states. Every server powering an AI application requires twenty critical metals, from gallium to scandium to tantalum to rhenium, metals most investors cannot name, let alone find on a balance sheet.
The military dimension compounds the problem further. Friedland describes the architecture of modern drone warfare, where AI-enabled targeting relies on low-Earth-orbit satellites with latency measured in millionths of a second. "If my satellite system has a millionth of a second less latency than yours," he says flatly, "you die, I don't." The Pentagon's old food pyramid of coal and oil has given way to a new one built entirely on critical metals. Drone systems, missile interceptors, hypersonic platforms, humanoid robots: all of it runs on the same supply chain that the world is rapidly failing to build.
The Vandals Took the Handle
The supply chain for mining equipment is itself in crisis. At Ivanhoe's Oyu Tolgoi mine in Mongolia, the ring gear for a primary mill motor carried a four-and-a-half-year delivery wait at the time of construction. "Today it'd be an 8 or 10-year wait," Friedland says, "if ever." Gas turbines needed to power data centers carry six- to eight-year backlogs, with force majeure clauses embedded in contracts citing unavailable critical metals.
When flooding struck one of Ivanhoe's Congo operations, Friedland needed specialized pumps capable of moving 1,300 liters per second from depths approaching the height of two Eiffel Towers. The Americans could not build them. The Germans could not build them. Only the Chinese could, and only because samarium-cobalt magnets for the electric motors remained accessible through Chinese supply chains. Invoking Bob Dylan, Friedland makes the point precisely: "The pump won't work because the vandals took the handle."
The Age of Hoarding
The geopolitical dimension is where the analysis turns most urgent. "We are in the low foothills, not the Rocky Mountains," Friedland says of the unfolding supply crisis. China has methodically acquired control of the critical materials supply chain while American capital flowed into Netflix, broadband, and semiconductor design. "We don't own these assets anymore 'cause all of our money went into Netflix," Friedland says. Mining companies in aggregate now represent less than 1% of the S&P 500, an all-time low, in part because net present value models designed for declining oil fields were misapplied to mines with century-long reserve lives.
In response, nations are transitioning from just-in-time supply chains to just-in-case strategies. "We're beginning the age of hoarding," Friedland says, noting that governments are quietly buying copper as a monetary proxy and not reporting it, making supply-demand modelling structurally unreliable. Russia and China have banned sulfuric acid exports, an essential reagent for roughly 25% of global copper production, driving prices from $150 per ton to over $1,000 per ton in eight months. The same acid is a feedstock for fertilizer, semiconductors, and food systems globally. "If we don't have more of it," Friedland warns, "food prices are gonna rise globally."
Innovation: A Closing Window
On whether technology can resolve the impasse, Friedland is careful. "We cannot find and responsibly mine the copper that we need for an energy transition with yesterday's technology. It's just impossible." His ventures, including Ivanhoe Electric and the newly formed iPulse, backed by tech investors and the governments of Japan, France, and the United States, are pursuing geothermal energy, cleaner extraction technology, and AI-assisted exploration. "With great difficulty," he concludes, it may be possible to innovate through the crisis. "But it's gonna be very tough, and it's gonna be a very closely fought thing."
Weisenthal and Alloway close the episode by noting, fairly, that solutions remain unclear. The constraint is not at one level of the supply chain. It is at every level simultaneously, from equipment manufacturing to critical input materials to permitting to capital allocation. Alloway captures it cleanly: "It's still not clear that the market will actually reward you for increasing production, because you still have the volatility and the supply problem." Before solutions can be built, the problem must be fully understood. Friedland's contribution to that understanding is substantial.
Five Key Takeaways for Advisors and Investors
1. The demand-supply gap in copper is historically without precedent. Maintaining 3% global GDP growth alone requires mining as much copper in the next eighteen years as humanity has mined in total across recorded history. Add electrification, AI infrastructure, and military modernization, and the gap widens further. Advisors should treat copper exposure not as cyclical commodity speculation but as a structural, multi-decade theme.
2. AI is a demand multiplier that most models have not yet priced. AI searches consume thirty times the electrical energy of a standard web query. As AI becomes ubiquitous across consumer and enterprise applications, the electricity and copper demand implications are compounding rapidly. Data center build-out, grid expansion, and cooling infrastructure all drive additional copper consumption that consensus demand forecasts have consistently underestimated.
3. Critical materials are now a primary geopolitical lever. China's control of the supply chains for gallium, scandium, dysprosium, and other critical metals gives it the ability to constrain Western AI development and military modernization. The US-China standoff over chips and critical metals is not a trade dispute. It is the defining strategic contest of the next decade. Portfolios with no exposure to critical materials producers carry a geopolitical blind spot.
4. Mining companies are structurally undervalued relative to their strategic importance. At less than 1% of S&P 500 market capitalization in aggregate, mining companies are priced as a legacy sector at precisely the moment their output becomes irreplaceable. The misapplication of oil-field NPV models to mines with decades of reserve life has suppressed valuations. As strategic buyers, hyperscalers, and sovereign governments begin competing for long-term supply, rerating of quality producers is a logical consequence.
5. Technological innovation is necessary but not sufficient on its own. The constraints on copper supply are geological, regulatory, geopolitical, and capital-structural. New technology from ventures like iPulse, Ivanhoe Electric, and others may improve the economics and environmental profile of extraction. But without meaningful reform of permitting timelines, a recalibration of capital toward critical materials, and a coherent national resource strategy, technology alone cannot close the gap. The Resolution Copper project in Arizona has been in permitting for 35 years. The window for acting ahead of the shortfall is closing.
1. Weisenthal, Joe, and Tracy Alloway. "Robert Friedland on the World's Monumental Shortage of Copper." Bloomberg Odd Lots, Bloomberg L.P., 12 Sept. 2026.
