Linguist Decodes Wolf Howls Using Silicon Valley AI Tools

Computational linguist Jeffrey Reed has begun separating individual wolf howls from background noise in Yellowstone National Park. He uses artificial intelligence to interpret the specific meaning of each sound. Reed holds a doctorate in ancient languages and spent 30 years in Silicon Valley. During that time, he built machines designed to replicate human language. He recently moved back to his hometown on the park's outskirts to apply those same technologies to wildlife.

Reed co-founded The Cry Wolf Project. The group is conducting the largest bioacoustics study of wild wolves currently in existence. He also co-founded Grizzly Systems. This company builds the hardware required for monitoring wildlife in the wild. His goal is to isolate, record, and interpret the vocalizations wolves make in their natural habitat. He aims to decode what specific howls signify to the pack.

Reed is currently writing a book about wolf communication. The manuscript is forthcoming from Little, Brown and Company. He previously worked on projects that allowed computers to mimic human speech patterns. That background informs his current work with bioacoustics. The project relies on software to filter out environmental sounds and focus on biological signals.

The segment aired as part of the TED Radio Hour episode titled Can curiosity save the wild? Phoebe Lett produced the piece. Sanaz Meshkinpour edited the audio. The episode explores how curiosity drives research in biodiversity. It also features related discussions on learning to communicate with sperm whales. Another segment covers the practice of imitating nature's voices through human vocalization.