The Critical Technology in Finding Critical Materials

Episode Summary

Executive Summary: The episode argues that critical minerals underpin electrification, AI, and modern infrastructure, but finding and developing them is slow, data-intensive, and increasingly strategic. Cobalt’s team explains how AI, hyperspectral imaging, digitized legacy records, and close geoscience/data-science collaboration improve exploration speed and drilling precision, especially for copper and lithium, while also reshaping mining’s environmental and geopolitical role.

Main Topics: Why critical minerals matter (Priority: 5/5): Copper, lithium, nickel, and cobalt are presented as indispensable inputs for EVs, batteries, solar, and data centers, with no real substitutes at scale for copper and lithium. The exploration challenge (Priority: 5/5): Deposits are concentrated by rare geological processes, making discovery difficult, expensive, and slow; exploration must happen years before demand peaks. AI and data in mineral discovery (Priority: 5/5): Cobalt uses satellite imagery, geophysics, geochemistry, and machine learning to narrow search areas, clean legacy datasets, and prioritize drill targets. Human expertise plus machine learning (Priority: 4/5): Geologists and data scientists work in tight collaboration, combining field judgment with algorithms to interpret rock, model uncertainty, and test multiple hypotheses. Drilling as information gathering (Priority: 4/5): Rather than only drilling for ore, the team drills to maximize knowledge and falsify hypotheses quickly, reducing wasted capital on million-dollar drill holes. Geopolitics and industrial strategy (Priority: 4/5): Mineral supply is tied to national interest, industrialization, and regional development, especially in Africa and along corridors like Lobito. The future of mining (Priority: 3/5): The industry is moving toward more predictive, precise, and technologically sophisticated mining, altering its public image and operational footprint.

Key Arguments: The energy transition requires vastly more metal supply, especially copper and lithium, so new mines must be discovered now rather than later. Copper and lithium are effectively irreplaceable at large scale because of their unique conductivity and battery performance. Most metal deposits are not randomly distributed; they form where geology has concentrated them, which makes targeted exploration essential. Modern exploration improves when disparate datasets are cleaned, digitized, translated, and combined into predictive models. AI helps by identifying patterns in continent-scale data, then refining targets down to camp-scale and drill-hole-scale decisions. Legacy paper maps and reports contain valuable information that can be unlocked through scanning, translation, and structuring. Drilling is expensive and should be treated as an information-optimization problem, not just a brute-force search. Multiple geological hypotheses should be tested simultaneously because uncertainty is inherent in subsurface exploration. Mining is behind oil and gas in some technologies, but methods like directional drilling and advanced imaging can be adapted. Critical minerals are increasingly defined not only by market price but by national security, industrial policy, and concentration of supply. Africa’s mineral projects can drive industrialization, skills development, and broader regional investment when done transparently and legally.

Data Points: EV share of global car sales: 14% - EVs already represent this share globally, reinforcing rising demand for battery metals. EV share of car sales in China: Majority in 2024 - China is cited as an example of rapid EV adoption driving mineral demand. Copper content in EVs vs gas vehicles: 4x more copper - EVs require significantly more copper than conventional internal combustion vehicles. World copper going to data centers by 2050: 6% to 7% - BHP estimate referenced for future copper demand from data centers. Time to find and develop a mine: Years to decades - Discovery, resource definition, mine construction, and extraction all take long lead times. EVs to be built for transition: About 2 billion - Estimated scale of vehicle buildout needed for the energy transition. New mines needed for EV supply chain: About 1,000 - Estimate for mines required to supply lithium, nickel, copper, and cobalt for EVs. Copper grade in major porphyry deposits: 0.5% to 0.6% - Typical lower-grade copper deposits discussed as benchmark examples. Copper grade in copper belt deposits: 2% to 3% - Average grade cited for copper belt deposits. High-grade Mangomba copper deposit: 5% to 6% - Example of an unusually high-grade discovery. Actual recovered product from mined rock: Less than 1% - After mining and processing losses, the usable copper fraction is very small. Project scale of airborne hyperspectral surveys: Thousands of square kilometers - Cobalt’s airborne imaging surveys are used to automatically interpret surface geology at scale. Core inventory at Mingomba: Nearly 100 kilometers - Quantity of rock core analyzed with machine-learning-assisted imaging. Model improvement in lithium exploration: Over an order of magnitude - Accuracy improved after incorporating true and false positives from field verification in Quebec. Drill hole cost: Up to $1 million - Single drill holes can be extremely expensive, driving the need for better targeting. Team growth in Zambia project: 10 rigs - Precision improvements over 24 months supported expanded drilling activity.

Pivotal Quotes: "if you can't grow it, you must mine it" — Connie Chan: A framing line for why mining is foundational to the physical economy. "You really can't tell what's underground until you drill." — Tom Hunt: Explains why exploration must use data to optimize expensive drilling decisions. "The best geologist is the one that's seen the most rocks." — George Gilchrist: Used to illustrate traditional geological expertise and the promise of machine learning at scale.

Implications: Critical minerals are becoming a strategic bottleneck for electrification, AI infrastructure, and national security. The winners will combine geology, AI, and data infrastructure to discover mines faster, drill smarter, and reduce environmental impact while building supply chains in geopolitically important regions.