AI Is Helping Researchers Decode Animal Calls, but Translation Is Unresolved

The appeal is not just curiosity. Con Slobodchikoff, founder of Zoolingua, a company working on dog-to-human speech, said decoding animal communication could offer a window into the inner lives of other species and help humans recognize that “animals are sentient creatures who think, who have hopes and dreams of their own.”

Teasing out meaning is difficult. To decipher howls, squeaks and clicks, researchers must match the sounds animals make with what is happening around them. Prairie dogs, for example, produce a different series of yips when a coyote approaches than when a hawk is nearby, according to the report. The work is painstaking, especially for species that are hard to observe, such as deep-diving whales or subterranean mole rats.

Animals also communicate through visual cues, touch, pheromones or combinations of signals, some of which humans may struggle to perceive. And without a multispecies Rosetta stone, researchers cannot easily know whether their translations are accurate.

Over the past 50 years, a growing arsenal of devices has made animal communication easier to observe. Hydrophones on the ocean floor can capture humpback whale song, and tiny wearable “biologgers” can record a bat’s every squeak, the report said. In the last decade, advances in AI have improved researchers’ ability to detect different calls, distinguish which sounds belong to which animal and spot patterns.

The Earth Species Project, a nonprofit founded in 2018 to decode animal communication, recently used machine learning to detect vocalizations of carrion crows and determine which bird made which call, according to MIT Technology Review. Researchers at the group suspect that more advanced AI could translate animals’ chatter.

Sara Keen, one of ESP’s behavioral ecologists, said large language models such as Claude and ChatGPT excel at finding relationships between words. ESP hopes its own NatureLM—trained on human speech, music and animal sounds—will be able to interpret the vocalizations of a variety of species. That promise, however, is still theoretical.

Yossi Yovel, a neuroecologist at Tel Aviv University, worries that such models will put too much emphasis on acoustic similarities and make mistakes, such as lumping together similar sounds that have different meanings. It is not clear when—or whether—a two-way conversation will happen.

Still, Slobodchikoff said many animals have an evolutionary structure similar enough that humans should be able to understand what they are saying. Researchers have amassed heaps of data on animal communication, but they are still trying to figure out how to decipher it all, the report said.