The language of whales may soon be decoded by AI and help humans chat with them

0

Scientists may not be very far from decoding the language of whales. An interdisciplinary group of scientists have started collecting data in order to use artificial intelligence (AI) to understand how whales communicate. The project was named Cetacean Translation Initiative (CETI). The first discussions on the possibility of decoding the sounds of sperm whales took place at Harvard University. An international group of scientists spent a year together at the Radcliffe Fellowship in 2017. Research and data collection began in full swing in 2020. If the project is successful, it would be the first time humans would understand the language of ‘another species. As a result, humans can also build a system to communicate with whales.

Shafi Goldwasser, director of the Simons Institute for the Theory of Computing at the University of California at Berkeley, noted a series of whale clicking sounds similar to Morse code, or the sound of a faulty electronic circuit. She presented the idea of ​​translating the language of whales through these clicks or “codas” to David Gruber, a marine biologist at the City University of New York. Next, Michael Bronstein, an Israeli computer scientist teaching at Imperial College London, considered a link between codas and natural language processing (NLP).

Biologist Shane Gero provided recordings of sperm whale codas from the Caribbean island of Dominica. Bronstein applied machine learning algorithms to this data. He told Hakai Magazine, “They seemed to work really well, at least with some relatively simple tasks.” But that was just a proof of concept.

Scientists and linguists still don’t know whether or not animals have a tongue. Animal utterances can only be called a language if they have semantics (vocalizations with fixed meanings), grammar (a fixed way of structuring sounds) and are not just innate sounds.

Whales generally dive in deep waters and communicate over great distances. Therefore, facial expression or body language does not affect their communication. Bronstein added, “It’s realistic to assume that whales’ communication is primarily acoustic. “

However, learning to decipher and communicate in the language of whales is also difficult for AI. The best-known AI language models are contained in GPT-3, which has a database of nearly 175 billion words. In comparison, the CETI database has less than 100,000 sperm whale codas. Scientists plan to expand the database to four billion codas.



Source link

Share.

About Author

Comments are closed.