The Project CETI team of researchers (the Cetacean Translation Initiative), is launching the largest ever “interspecies communication” project. What is the target species? The target species? While the task will be challenging, researchers will have AI, drones, and robotic fish to help them.
National Geographic has recently updated readers about the latest progress reports. David Gruber, National Geographic Explorer and marine biologist discussed the project with National Geographic. He says that the project is ultimately about helping people connect with nature and sea life.
In Depth Project Ceti AI Dominica Welch
Gruber explained to National Geographic, “I had the idea that if people could fall in Love with jellyfish, then they could fall in LOVE with everything.” “But there is something about whales which really taps into our human curiosity.”
Gruber began to study whale language. A computer scientist colleague heard Gruber’s clicks. He said they reminded him of Morse code, an alphabet where letters are represented using combinations of light signals and sounds of different lengths. Gruber joined his colleague. They were joined by several of their fellow researchers who specialize in artificial intelligence.
Also, read Eic Dan Roth Fischer Axios
Gruber and his colleagues want to greatly increase the number of whale conversations. To train machine-learning algorithms–algorithms that can glean relevant patterns from datasets–, the researchers need to go from small data to big data. Gruber claims that there are only a few thousand examples currently of whale communication. The algorithms will only work if there are millions of examples.
The CETI team has launched a multi-pronged effort to collect large amounts of whale conversations. For example, the team will place microphones on the ocean floor close to popular sperm whale locations. Soft robotic fish will also be deployed to assist the whales and capture their communications. Researchers will use drones to drop microphones in the water.
Gruber predicts that all of these language examples could be used to create the largest animal behavioral dataset. The algorithms will be able to decipher whale language and allow researchers to communicate with the whales, mainly to determine if they respond predictably. If they do, it will confirm our understanding of basic language elements. It will remain a mystery as to what the whales, with their enormous brains, think about the world.