Could We Speak to Dolphins? A Promising LLM Makes That a Possibility

Science Quickly

Scientific American

Science

4.2 • 639 Ratings

🗓️ 23 May 2025

⏱️ 18 minutes

🔗️ Recording | iTunes | RSS

🧾️ Download transcript

Summary

Dolphins have a broad vocabulary. They vocalize with whistles, clicks and “burst pulses.”This varied communication makes it challenging for scientists to decode dolphin speech. Artificial intelligence can help researchers process audio and find the slight patterns that human ears may not be able to identify. Reporter Melissa Hobson took a look at DolphinGemma, a large language model created by Google in collaboration with the Wild Dolphin Project and the Georgia Institute of Technology. The project seeks to unravel the clicks from the whistles and to understand what dolphins chat about under the waves. Recommended reading: Read our article about DolphinGemma: https://www.scientificamerican.com/article/can-ai-let-us-chat-with-dolphins/ Watch our video about the project: https://www.tiktok.com/@scientificamerican/video/7499862659072871723 Keep up with Hobson’s reporting: http://www.melissahobson.co.uk/ Tell us what you think! Take our survey for the chance to win some SciAm swag! http://sciencequickly.com/survey E-mail us at [email protected] if you have any questions, comments or ideas for stories we should cover! Discover something new every day: subscribe to Scientific American and sign up for Today in Science, our daily newsletter. Science Quickly is produced by Rachel Feltman, Fonda Mwangi, Kelso Harper, Naeem Amarsy and Jeff DelViscio. This episode was co-hosted by Rachel Feltman. Our show is edited by Alex Sugiura with fact-checking by Shayna Posses and Aaron Shattuck. The theme music was composed by Dominic Smith. Learn more about your ad choices. Visit megaphone.fm/adchoices

Transcript

Click on a timestamp to play from that location

0:00.0	Understanding the human body is a team effort. That's where the Yachtel group comes in.
0:05.8	Researchers at Yachtolt have been delving into the secrets of probiotics for 90 years.
0:11.0	Yachtold also partners with nature portfolio to advance gut microbiome science through the global grants for gut health, an investigator-led research program.
0:20.1	To learn more about Yachtolt, visit yawcult.co.
0:22.7	.jp. That's Y-A-K-U-L-T.C-O.JP. When it comes to a guide for your gut, I'm Rachel Feldman.
0:59.0	There are a few animals that pretty much everyone likes.
1:02.6	Fluffy pandas, cute kittens, real tigers.
1:06.0	Dolphins would probably make the list for most folks, too.
1:11.6	They're intelligent, playful, and have those permanent goofy grins on their faces. Watching them dart around in the water kind of makes you wonder, what are those guys thinking?
1:16.6	It's a question many scientists have asked.
1:18.6	But could we actually find out?
1:21.6	And what if we could even talk back?
1:23.6	Freelance Ocean Writer Melissa Hobson has been looking into a new project that's been making a big splash, sorry, in the media.
1:31.7	It's being billed as the first large language model, or LLM, for dolphin vocalizations.
1:38.3	Could this new tech allow us to actually communicate with dolphins?
1:42.7	Here's Melissa to share what she's learned.
1:55.7	When you dip your head under the waves at the beach,
1:58.3	the water muffles the noise around you,
2:00.2	and everything goes quiet for a moment. People often assume that means the ocean is silent, but that's really not true.
2:06.6	Underwater habitats are actually full of noise. In fact, some marine animals rely heavily on sound for communication,
2:14.6	like dolphins.
2:22.3	If you've ever been in the water with dolphins or watch them on TV, you'll notice that they're always chattering, chirping,
	...

Please login to see the full transcript.

Previous episode | Next episode

Disclaimer: The podcast and artwork embedded on this page are from Scientific American, and are the property of its owner and not affiliated with or endorsed by Tapesearch.

Generated transcripts are the property of Scientific American and are distributed freely under the Fair Use doctrine. Transcripts generated by Tapesearch are not guaranteed to be accurate.