December 15, 2020
Reading time: 1 minute
“Climate change is not an on/off switch,” says Dr. David Rolnick, the co-founder and chair of Climate Change AI, an initiative catalyzing impactful work at the intersection of machine learning and climate change. “But we have the ability to decide exactly how bad it will be based on our actions now.”
Dr. Rolnick presented key insights from his group’s paper Tackling Climate Change with Machine Learning at a December speaker event hosted by the Leadership in Environmental and Digital innovation for Sustainability (LEADS) program and Sustainability in the Digital Age. After a brief introduction of Machine Learning (ML), defined as groups of techniques that automatically extract patterns from large amounts of data, he presented some areas where ML can really scale up mitigation and adaptation efforts. These include remote sensing, improving efficiency, forecasting, accelerating simulations, and accelerating experimentation. At the close of his talk, he reiterated that machine learning “is not a silver bullet, it is just one tool for fighting climate change.”
As a follow-up discussant, Dr. Catherine Potvin, a Canada Research Chair in Climate Change Mitigation, echoed this sentiment, emphasizing how critical it is to reflect on the purpose served by digital innovations, and that solutions should be defined by society and based on evidence.
This kicked off a rich discussion from a virtual audience of about 50 students, professors and other interested individuals. Attendees commented on the necessity of cross-disciplinary collaboration, that ML applications cannot be developed in isolation from climate science but must instead be built together with stakeholders cognizant of the community, policy and engineering implications. Questions were also posed on how to stimulate computer science literacy among students and policymakers, and on how to develop curricula and training so that basic ML concepts are demystified for a wider population. As one participant commented, “Literacy makes it less possible to misuse something.”
On this theme, the LEADS and Sustainability in the Digital Age speaker series will offer training sessions in 2021 to provide more hands-on practice and in-depth sharing in the future sessions. Visit the LEADS website for current activities.