Talking Beasts?: Exploring the Role of AI in Human-Animal Conversations
Can animals talk? This is a question that has perplexed biologists and linguists for centuries. Even as animals have ways of communicating with each other, including using auditory, tactile, or chemical signals, their communication is often simplistic. As such, it pales in comparison to the layered and sophisticated ways of human communication. This is because they have developed many ways of communication: language, gestures, writing, sound, visualization.
Nonetheless, the emergence of artificial intelligence technologies has raised the prospects of humans finally being able to unveil the mysteries of communication between animals, and even developing modalities to facilitate human-animal communication.
For example, a behavioral ecologist at the University of St Andrews named Christian Rutz has found out that the new Caledonian crow is not only capable of manufacturing tools, which for the longest time was thought to be a skill exclusive to humans, but this species of the crow is also a perfectionist, that strives to work until it is absolutely satisfied with its tool.
It is precisely in facilitating discoveries like these that artificial intelligence can provide us with fascinating insights. For example, the Earth Species Project is a non-profit organization made of AI and data scientists, biologists, as well as conversation experts, which is trying to decipher animal communication such as crow calls using machine learning tools.
Projects like these attempt to decode animal vocalizations through training AI on colossal data sets available to biologists. For example, work done by Shane Gero at Carleton University in Ottawa has led to the revelation that sperm whales use specific sound patterns called codas to identify one another and communicate. Sperm whales learn to use codas much in the same way that toddlers learn words: repeating the sounds made by adults around them. Using AI technologies to translate these codas through feeding data to a neural network, Gero was able to correctly identify a small subset of individual whales from the codas used to refer to them. Currently, Gero’s team is using AI to devise ways of communicating with whales.
The hope underlying several such projects is that this would allow us to augment our conservation and welfare efforts by better understanding the needs and responses of animals. These technologies are bound to have an implosive effect on the anthropomorphic assumptions that structure our lives by refuting the idea that humans are automatically superior to other species.
Sources