Google’s artificial intelligence technology is being adapted to help researchers potentially communicate with dolphins through a new open AI model, the company said on Monday.
The model, called DolphinGemma, is a collaboration between Google and the Wild Dolphin Project (WDP).
The project is building on more than 40 years of research into dolphins and their communication, trained on an acoustic database of Atlantic spotted dolphins created by the WDP, which associates correlates sound types with behavioural contexts.
These include as signature whistles to help calves find their mothers, burst-pulse “squawks” that are observed during fights, and click “buzzes” that are often used during courtship or chasing sharks.
Coffee drinkers face new price hikes and the latest Trump tariff twists
If planning laws were changed, obsolete offices could be converted into housing to ease Dublin’s rental crisis
Johnny Lappin: ‘I got scammed by a rogue so-called roofing contractor. I foolishly paid him in cash’
The silence of the CEOs in the face of Donald Trump’s tariff chaos
Using this information as a training set, DolphinGemma can pick out patterns in communications between the mammals more quickly than human analysis alone.
“This approach in the quest for interspecies communication pushes the boundaries of AI and our potential connection with the marine world,” Google said in a blog post.
“By identifying recurring sound patterns, clusters and reliable sequences, the model can help researchers uncover hidden structures and potential meanings within the dolphins’ natural communication – a task previously requiring immense human effort.”
“Eventually, these patterns, augmented with synthetic sounds created by the researchers to refer to objects with which the dolphins like to play, may establish a shared vocabulary with the dolphins for interactive communication.”
The Google-developed system also uses the company’s audio technologies such as the SoundStream Tokenizer to generate dolphin-like sound sequences, and processes it with technology that can run directly on Pixel phones.
Scientists will be able to adapt and use the model in their own projects.
The use of Pixel phones eliminates the need for custom equipment, while also lowers power consumption and shrinks the device’s cost and size, bringing a key advantage to field research.
