Back

Speaker "Robert Bernard" Details Back

 

Topic

AI and Climate Change: Opportunities and Pitfalls

Abstract

In recent years throughout a myriad of professional fields, ESG (environment, social, and governance) has increasingly become more prominent and the existential threat of climate change has become even more pervasive and urgent.  In parallel, AI (artificial intelligence) has steadily been making progress, especially in the fields of deep learning, as applied to language understanding and generation as well as image processing and automated interpretation.  Indeed, AI (and data science and machine learning) can certainly play a role in helping to mitigate the climate crisis, through improved forecasting, optimization, and scheduling.  However, the rise of these complex deep learning models also contributes significantly to the climate crisis, through the energy resources required to train them on massively parallel computers.  In this paper, I will review the opportunities that AI can play in helping manage climate change, but also detail the pitfalls that may increase if its growth is unchecked.  I will conclude by offering practical suggestions on how individual companies can use AI to improve company performance, satisfy regulatory requirements, and help the Earth along the way.

Profile

Rob Bernard is the Director of Climate Risk Modeling at PwC where he focuses on physical risk modeling, climate change, and related risk management and risk mitigation strategies. Rob leads a team that develops analytics, models, and quantitative frameworks that estimate the deleterious effects of climate change.
Rob has developed analytics and models in a variety of industries, from simulating long-term land use and demographic change for small towns, to creation of machine learning models of fraud detection, to creation of innovative natural language processing techniques for measuring engagement, commitment, and understanding in online discussion forums.
He has a Bachelor's degree in Cognitive Psychology from Princeton, a Master's degree in Urban and Regional Planning and Gaming/Simulation studies from Michigan, and he is currently completing his Master's thesis in Predictive Analytics at Northwestern.