Applying Artificial Intelligence for a Sustainable Energy Future
Jef Caers, Professor of Geological Sciences, Stanford University
Director, Stanford Center for Earth Resource Forecasting
We are seeking a postdoctoral researcher for a 2-year position to research the application of artificial intelligence for the exploration of critical battery metals. The candidate will collaborate directly with KoBold Metals (KoBold), a San Francisco Bay Area company backed by Bill Gates, Jeff Bezos, and Andreessen Horowitz, that is developing artificial intelligence to improve the efficacy, efficiency, and sustainability of critical mineral exploration and development.
The project will focus on sequential decision making under uncertainty using the latest research in intelligence systems and agents as well as predictive modeling based on uncertainty quantification with sparse and fuzzy data.
Summary of research project
Battery metals are to the renewable energy and electric vehicle revolution what iron was to the iron-age and what coal was to the industrial revolution. As electric vehicles and renewable electricity become widespread, many new mines will be needed; but before new deposits can be mined, they have to be discovered. In collaboration with the Stanford Center for Earth Resources Forecasting, KoBold is building AI systems to find new ore deposits rich in critical battery materials by combining the world’s top mineral explorers with an outstanding team of data scientists and software engineers.
Scientific discovery works through a series of experiments with Bayesian updating. In this manner, mineral exploration requires careful planning of sequential data acquisition over large areas to fill in sparse datasets. This project will develop an artificial intelligent agent termed the Intelligent Prospector that assists humans in the optimal planning of geophysical and geochemical data acquisitions campaigns. The intelligent agent will interact with KoBold's Machine Prospector technology, which has been developed to provide quantified uncertainty from existing data sets. We envision the Intelligent Prospector will use the latest developments in AI such as Partially Observed Markov Decision processes, reinforcement learning, and Monte Carlo Tree Search Methods to accelerate discovery and thereby reach the ultimate goal of a battery-powered sustainable energy future
PhD in any STEM field
The candidate will also need experience with databases and Python scripting. We are looking for candidates with:
1. Demonstrated computer science expertise in artificial intelligence, data science programming, big data manipulation, and cloud computing
2. Having worked on real data with quantified uncertainty
3. Research as demonstrated through publications in scientific journals
Send a CV, statement, and a list of three referees
In your statement please elaborate explicitly on the following questions
● What expertise in AI do you have that can contribute to developing intelligent systems?
● What is your level of python programming as provided through examples?
● What are real data cases and problems that you have worked on?