Given the latest advances in Artificial Intelligence (AI), humanity is on the way to decoding animal language. It could revolutionize the way we live with animals.
Deciphering the animal language system is an ambition that dates back a few years, with advances in machine learning and deep learning. But we know that talking with animals is one of the fantasies that have always inspired humans the most. Numerous literary, cinematographic and artistic works attest to this fact, such as the story of Marcelino Bread and Winethe newborn child abandoned at the gates of a Franciscan monastery. Accompanied by his faithful dog, this boy helps the animals to manage their problems, being the only human to really understand the context.
At present, serious projects are being carried out on the subject and some results are particularly promising. If civilizations and other communication systems really existed under our noses and that we realize it, this can lead to great upheavals. It might cause the human to reconsider its models, its system of understanding and its relationship current with animals.
Researchers train an AI to understand the language of several animals
In a recent study documented in the New York Timesfive teams of researchers analyzed moans of different animals. To do this, the experts used machine learning algorithms. In effect, they train Artificial Intelligence to understand language rodents, lemurs, whales, chickens, pigs, bats, cats, dogs, etc.
It should be emphasized that analyzing animal language is a matter ofa bigger challenge that of understanding and translating the human language system for the machine. Artificial Intelligence must be trained to know what to look for and how to organize work and data. For this, it will be necessary collect an almost incalculable amount of dataincluding voice recording systems.
But still, this is not enough, because it is also necessary that these voice recordings correspond with the visual social behaviors of animalsincluding pictures. In this regard, however, we have already witnessed the creation of an application to identify a bird near them thanks to descriptions or a photo. This is Cornell’s Merlin technology.
DeepSqueak AI Shows Spectacular Prowess
In another research, published in the the wall street journalresearchers have demonstrated the amazing capabilities of Artificial Intelligence in the field. Indeed, some algorithms like DeepSqueak are able to spot subtle patterns that might escape human listeners. Concretely, this Artificial Intelligence is already capable of distinguishing individual sounds of animals from those that they emit in different contexts.
She can also break their vocalizations down into smaller parts. More impressive, the machine knows how to qualify the cries emitted by the animals when they feel good and when they feel bad. These are crucial steps to decipher their meaning, and towards real human-animal communication.
SEE ALSO: An artificial intelligence recognized as the owner of a patent
CETI: sperm whales could be the key
One of the most encouraging research on this is project CETI (Cetacean Translation Initiative). Together, machine learning experts, marine biologists, roboticists, linguists and cryptographers are trying to decode the communication of sperm whales.
But why sperm whales in particular?
As a reminder, this species has the biggest brain of all. Additionally, sperm whales display similar characteristics to humans. Indeed, studies have shown that sperm whales have a conscious thought. They would also be able of planning, speaking and feeling. In addition, they live in groups of 20 to 40 individuals. The members are linked by strong family ties.
The researchers believe that their sophisticated acoustic communications provide an excellent basis for collect data. Ultimately, the resulting technology could be applied to other animals. It will be seen as a tool for make better decisions and take the right actions.
Fascinating discoveries are fast approaching
The CETI team recognized that sperm whales might not be so fascinating than we thought. But that doesn’t seem to discourage her from doing the job.
“But we don’t think that’s the case. In my experience as a biologist, whenever I’ve really looked at something up close, there’s never been a time when I’ve been disappointed in animals. »
David Gruber, Research Manager
Project experts plan to make Artificial Intelligence publicly available so that it can be quickly enriched with data. In a long-term view, the goal is to be able to apply it to decipher animal language as a whole.
“If we discover that there is an entire civilization right under our noses, it could lead to a change in the way we deal with our environment. And maybe that will translate into a greater respect for the living world. »
Michael Bronstein, machine learning lead for the CETI project
The articles published on the subjects are encouraging. But we have to recognize that there is still a long way to go to build an exhaustive database. This must bring together all animal communication systems and networks. The machines will thus be more likely to make the right decisions, and recommend the appropriate actions depending on the context.