Artificial intelligence is being used to talk to animals

Ali Hussain

It may soon become possible to use bioacoustics and artificial intelligence to communicate with animals, according to Dr. Karen Bakker, a professor at the University of British Columbia and a fellow at the Harvard Radcliffe Institute for Advanced Study.

The scientific community has long been interested in interspecies communication, with the most successful attempt being Koko, the gorilla who managed to use American Sign Language. Critics of Koko and other animals that have since been taught sign language claimed that this method is not true communication.

Humans “should have been thinking about [the animals’] abilities to engage in complex communication on their own terms, in their own embodied way, in their own worldview,” Bakker said in an interview with Scientific American.

Bakker recently published a book, titled “The Sounds of Life, How Digital Technology is Bringing us Closer to the World of Animals and Plants,” presenting a possible solution to this conundrum.

The complicated communication systems that animals use worldwide are incredibly difficult to replicate, let alone understand. Over 8.7 million species of plants and animals exist, each with its own communication system used to indicate food and tools necessary for survival and passing on offspring.

Many of these species communicate at audio levels that humans cannot process. Bats use ultrasound at frequencies of 9 to 200 kilohertz, allowing them to bounce messages to each other. Elephants use frequencies as low as 1 to 20 hertz, which can be heard up to 10 kilometers away by other elephants. Neither of these frequencies can be heard by the human ear, which has a range of about 20 hertz to 20 kilohertz.

Fortunately with the advantage of technology, humans can now use a combination of bioacoustics and AI to better understand nonhuman species’ communication.

“You and I could never sing like a whale or buzz like a bee, but computers and biomimetic robots can,” Bakker wrote. “Our digital devices have brought us to the brink of a new era in digitally mediated interspecies communication.”

To digitally communicate with animals, incredibly small digital records are attached to animals deep in the ocean and high in the sky. These recorders remain attached to the animal around the clock, allowing scientists to receive audio information free from human interference.

The records obtain an incredible volume of audio information that would take many lifetimes to process. This is where AI comes in.

“The same natural language processing algorithms that we are using to such great effect in tools such as Google Translate can also be used to detect patterns in nonhuman communication,” said Bakker.

In her book, Bakker described a success story of AI’s analysis of speech patterns in a population of Egyptian fruit bats, led by Dr. Yossi Yovel of Tel Aviv University.

After recording two dozen bats for almost three months, the AI was able to illustrate adequate knowledge of the complexity of bat’s speech. Bats distinguish between genders when communicating; they have “signature calls” that function similarly to the names humans use.

With humans, a mother will tend to speak louder when talking to her child. These mother bats did the opposite, lowering their voices when talking to their child, eliciting a babble response.

Bees communicate with other bees by dancing. Bakker highlighted a researcher named Dr. Tim Landgraf of Freie Universität Berlin who was able to encode this dancing system into a “RoboBee” that could enter the hive, communicate with other bees and lead them to specific flowers to pollinate. The hope of this experiment was to use the information so that honeybees would be led away from polluted lands and toward better nectar.

Bakker hopes that these breakthroughs in interspecies communication may lead to a better understanding of humanity’s place on earth, and that the newfound conversation with nature may aid conservation efforts.