Summary
David Rolnick, a scientist and pioneer of machine learning in the fight against climate change, founded Climate Change AI. This group brings together experts and stakeholders interested in the intersection of climate change and machine learning. The group provides a call to arms from the machine learning community to describe the opportunities that exist and showcase the terrific work that had already been done in this space. The group is focused on catalyzing impactful work at the intersection of climate change and machine learning. The group runs events, workshops, resources for learning, and a multi-million dollar grants program to fund work in this area and catalyze the creation of new datasets.
In his paper, "Tackling Climate Change with Machine Learning," David and his co-authors highlighted a wide range of opportunities in various areas, such as distilling large unstructured datasets into useful information, optimizing complicated systems, forecasting, and scientific modeling and discovery. The group also works on fast approximations to radiative transfer and physics-based constraints, which allow for speeding up simulations in contexts like atmospheric physics or aerodynamics of vehicles. Machine learning can push the envelope of methodological boundaries in uncertainty quantification and incorporate physics-based constraints in neural network predictions.
While AI is a powerful tool to have an impact on climate change, it is not a silver bullet and needs to be used in collaboration with different people and experts in machine learning and relevant application areas. Collaboration is essential to avoid pitfalls and ensure a pathway to meaningful impact. Specific knowledge is generally essential since there are always constraints and contextual information that aren't just captured in the data. It is crucial to consider how something is going to be used to build in any deployment considerations right from the start.
The speaker advises students to become well-versed in a specific area and to become an expert in machine learning and power systems or ml and infrastructure. The most impactful areas of exploration involve integrating existing data with domain knowledge and with other considerations which are not nicely packaged in the way that we're used to in the context of pure machine learning. The choices that technologists make can change the impact of new technology on climate change, both implicitly and explicitly.
David Rolnick is optimistic that we can make things less bad in terms of climate change. Humanity has already seriously impacted the world in many ways that are if not irreversible then certainly within the time span of society, but we get to choose how bad things become or how good things can be as well because many of the things we have done have also made life better for many people. The technologies exist to use renewable and low-carbon energy and machine learning can help boost adoption or mitigate bottlenecks to adoption of these technologies.
