Can AI help discover new copper deposits?

Critical minerals like copper and lithium underpin modern technology, from electric vehicles to AI data centres. To hit global net zero targets, the International Energy Agency recently estimated that 194 new copper mines will be needed by 2050. But could machine learning help solve its own resource problem? That’s the pitch from KoBold Metals, a Silicon Valley startup which has already pulled in more than $1 billion in investment including Bill Gates and Jeff Bezos. By using computer simulations based on a vast geological database, KoBold says it can find new seams of critical minerals, faster. Its founders – Kurt House and Josh Goldman – have been speaking to the NYT and WSJ about their hunt for copper in Zambia’s Copperbelt province. Earlier this month, they informed investors they had made potentially the largest copper discovery in over a decade. The company is planning to go beyond mineral exploration and invest $2.3 billion into setting up its own mining operations. The US government, very aware of its current reliance on China for critical mineral supply, is expected to finance a new railway to help export the copper.