Biodiversity & AI: Hype Vs. Reality
Last week I sat down with Giada Ferraglioni from Carbon Pulse to sift through the noise I'm seeing on #biodiversity and #AI 👇
My view is based on years of working with machine learning and LLM in enterprise tech and, in the last few years, working with the Nature Tech Collective members on their nature data.
My work has involved #eDNA, #bioacoustics, smart camera traps, smart soil sticks, #LiDAR drones, and satellite image data sets. With nature scientists and companies at the forefront of nature tech, we've explored IBAT, the Ecosystem Integrity Index, the Biodiversity Intactness Index, and other models.
Last year, I had the honor of contributing to the #TNFD scoping study for a public nature data utility tool (link in the comments 👇). Together, we built an MVP of a nature data library, a repository of over 800 datasets meticulously curated across land, ocean, freshwater, and atmosphere. Each dataset is a valuable piece of the puzzle, contributing to our collective understanding of nature.
I must say that there are no shortcuts when it comes to harnessing the power of machine learning and LLM models to support decision-making in nature conservation and restoration.
WE HAVE TO COLLECT THE DOTS BEFORE WE CAN CONNECT THE DOTS
We need more clean, organized, and indexed biodiversity data across all realms: land, ocean, and freshwater. On species, on genetics, on habitats
It's crucial to emphasize that understanding biodiversity is not a task that can be accomplished from space. We must invest significantly in biodiversity data infrastructure, train IPLCs, and provide data bounties to collect ground truth data. This includes tagging bird recordings, expanding DNA databases, and sampling more soil. Only then can we truly comprehend and conserve our natural world.
It's also clear that we need to foster more multidisciplinary collaboration. Many conservation science teams working in space lack a deep understanding of modern data operations and data stacks. I foresee a new profession emerging: the nature data engineer, a role that will bridge the gap between conservation science and modern data operations.
Are you working with biodiversity data? Reach out or leave your thoughts after reading this article 👇
https://carbon-pulse.com/293270/
#naturedata #natureAI #naturerisk #biodiversity #biodiversityloss #naturetech
