The A16Z podcast episode features a conversation with Connie Chan, a General Partner at A16Z, and three experts from Cobalt, a mineral exploration company using AI and human intelligence to discover critical materials: Tom Hunt, VP of Technology; Nphikei Makai, mining and civil engineer; and George Gilchrist, VP of GS Sciences. The discussion centers on the increasing demand for critical metals driven by the energy transition, the challenges of finding and extracting these materials, and how technology is revolutionizing mineral exploration.
Connie Chan emphasizes the fundamental role of mining in modern life, stating, "If you can't grow it, you must mine it." She highlights the growing demand for metals like copper, lithium, and nickel, crucial for electric vehicles, data centers, and other emerging technologies. The urgency stems from the lengthy process involved in discovering, developing, and operating a mine, which can take decades. Given the projected supply gap in the coming decades, finding new metal deposits needs to happen now.
The Cobalt team delves into the complexities of metal exploration. They emphasize that metals, while abundant in the Earth's crust, are difficult to find in concentrated deposits that can be cost-effectively and sustainably extracted. The geological conditions necessary for concentrating metals are unique and vary depending on the metal sought. There is no single "formula" for making a discovery, making exploration challenging.
George Gilchrist notes that discovering new mines is becoming increasingly difficult. Many surface deposits have already been found, necessitating the use of more advanced tools and data to explore beneath the surface. Historically, smaller companies willing to take risks and explore unconventional areas have often made discoveries. Exploration is the new horizon to expand what is needed to power technology.
Tom Hunt elaborates on how AI is transforming exploration. He emphasizes the importance of integrating diverse data sources, from satellite imagery to geochemical data, to identify promising areas. AI algorithms are used for image recognition, classification, and data cleaning to extract insights from these large datasets. The ultimate goal is to optimize drilling, which can be extremely expensive, by pinpointing the most prospective locations.
The discussion highlights the importance of high-quality, structured data. Much of the historical data is unstructured, existing in handwritten records, paper archives, and diverse formats. Cobalt is investing in digitizing these records, translating them when necessary, and extracting structured data that can be used by AI algorithms.
A key element of Cobalt's approach is the collaboration between geoscientists and data scientists. These teams work together to interpret data, build models, and guide exploration efforts. Nphikei Makai emphasizes the importance of creating a shared language and understanding between these disciplines, even coining new words to bridge the gap between the two areas.
Cobalt is also innovating on the hardware front. They are utilizing hyperspectral imaging on airborne systems to capture detailed chemical information about the Earth's surface. This data, combined with ground truth data from geologists, allows for the creation of data-driven geological maps.
George Gilchrist provides examples of how AI has surprised him. He notes that AI algorithms can identify relationships between elements in incomplete datasets, allowing Cobalt to utilize data that would otherwise be unusable. He also emphasizes the efficiency gains from digitizing maps, enabling geologists to quickly access and interrogate information.
The conversation covers prioritizing information in a data-rich world. The geoscientists' experience in specific environments is critical, and the collaboration between geoscientists and data scientists is pivotal to success. It also notes that it is data driven decision making that uses testing to get the most information possible out of each hole or area that is drilled into and evaluated.
The discussion then shifts to the future of mining. The team notes that they want to change the image of mining and make it a more tech forward and environmentally sustainable undertaking by leveraging data driven decisions with information from machine learning to reduce the need for mining, waste and resources.
Finally, the guests cover critical minerals, highlighting copper and lithium as the most crucial for electrification. Rare earth elements and minerals with concentrated supply chains also play a role in the importance of critical minerals. They also touch on the opportunities for countries like Zambia, where Cobalt's investments are stimulating industrialization and economic growth. The team has worked very hard and with great strides to change the image of mining in the country. The podcast ends on a note of optimism, emphasizing the importance of applying new technologies to solve the challenges of mineral exploration and ensure a sustainable supply of critical materials.