An AI-equipped underwater robot is tracking sperm whale clicks in real time to decode the animals' complex communication system [1, 2].

The project represents a significant leap in marine biology by using artificial intelligence to translate "codas," the rhythmic click sounds whales use to communicate. If successful, the effort could allow humans to understand the sophisticated social structures, and meanings behind deep-sea vocalizations [1, 2].

Developed by David Gruber, the founder of Project Cetacean, the robot is a specialized underwater glider designed to operate in the dark depths of the ocean [1, 2]. Unlike traditional research vessels, the glider can autonomously navigate the deep sea to remain in close proximity to whale pods [1].

Gruber compared the technology to autonomous vehicles to explain its capabilities. "This robot is similar to 'Waymo,' the self-driving car," Gruber said. "It makes its own decisions on how to move underwater and, while slow, it can stay near whales very efficiently for several months [1]."

This endurance allows the research team to secure long-term data sets that were previously impossible to collect [1]. By recording these sounds in their natural environment over extended periods, the AI can better identify patterns, and linguistic structures within the whales' clicks [1, 2].

The robot's ability to function independently in the deep ocean reduces the need for constant human supervision and minimizes the disruption of the whales' natural behavior [1, 2]. This approach enables a more organic collection of acoustic data, which is essential for the AI to accurately map the relationship between specific sounds and the whales' actions or social interactions [1].

This robot is similar to 'Waymo,' the self-driving car.

The integration of autonomous AI gliders into marine research shifts the study of cetacean communication from passive listening to active, long-term tracking. By mirroring the mobility of the whales, researchers can capture a more comprehensive linguistic database, potentially bridging the gap between human technology and non-human intelligence.