The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence) Visualizing Climate Impact with GANs w/ Sasha Luccioni - #413
6 snips
Sep 28, 2020 
Sasha Luccioni, a Postdoctoral Researcher at the MILA Institute, delves into the intersection of machine learning and climate change. She discusses how Generative Adversarial Networks (GANs) are being utilized to visualize climate impacts, shedding light on the technical challenges involved. The conversation also touches on the environmental footprint of AI, emphasizing the need for sustainable practices and collaboration in the field. Sasha aims to encourage effective communication around climate issues, making complex data accessible for public understanding.
AI Snips
Chapters
Transcript
Episode notes
Connecting AI and Activism
- After her PhD, Sasha Luccioni realized her AI work didn't align with her climate activism.
- She convinced Yoshua Bengio to hire her by creating a climate visualization GAN demo.
Early Activism
- Sasha Luccioni's climate activism began with vegetarianism at age 12 and deepened after having children.
- She reduced flights, embraced composting, and joined local environmental groups.
Localizing Climate Change
- Visualizing climate change impacts locally helps make the abstract threat more concrete and personal.
- It brings the consequences of far-off events like floods or fires closer to home.