Imagine walking through the woods with an app that can accurately identify, organize, and monitor the biodiversity of that ecosystem via vocalizations being made by birds and other animals. This project by Google’s A.I. Experiments may be a step in that direction. Using bird calls from the Macaulay Library’s Essential Set for North America, the team collaborated with the Cornell Lab of Ornithology to organize thousands of bird sounds using machine learning or artificial intelligence. In other words, the computer organized the short bird call clips without any programming by using an algorithm called t-SNE.
Browse the resulting AIExperiments interface with Google. Related DIY: A list of almost 100 common feeder birds and what they like to eat.More from the Cornell Lab and more from Google. Plus, Karl Sims’ Evolved Virtual Creatures (1994) and Can A Thousand Tiny Swarming Robots Outsmart Nature?