The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence) AI for Ecology and Ecosystem Preservation with Bryan Carstens - #449
Jan 21, 2021
Join Bryan Carstens, a professor at The Ohio State University and leader in biodiversity research, as he dives into the fascinating intersection of machine learning and ecology. Discover how his lab utilizes ML to tackle species formation, uncover genetic insights, and predict conservation risks amidst a looming extinction crisis. Carstens discusses the shift from traditional data collection to digitized methods, the complexities of ecological modeling, and the future of interdisciplinary collaboration in understanding our planet's critical biodiversity.
AI Snips
Chapters
Transcript
Episode notes
Hidden Biodiversity
- Biodiversity is less understood than commonly assumed, with only 1-10% of species described.
- Most new species are discovered in labs and museums, not unexplored jungles, by reclassifying existing specimens.
Misclassified Species
- Many undescribed species exist within known groups, like mammals, due to misclassification.
- Predictive modeling helps identify where these unknown species might be, guiding research efforts.
Predicting Plant Conservation Risk
- Carstens' lab created a model predicting plant conservation risk with 85% accuracy using trait, environmental, and geographic data.
- This model helps prioritize conservation efforts by identifying at-risk species needing further investigation.