Stunning, world-changing news has been trying to click and trill its way through the chaotic cloud of media that serves as humanity’s proxy for global collective consciousness. Did you miss it? Don’t worry, I will fill you in.

There have been some major breakthroughs in our efforts to learn to talk to whales. There is growing scientific proof that sperm whales have what must be called a language. Not just pretty musical sounds, or simple “let’s go eat” signals, but an actual language, with all the combinatorial possibilities for information exchange and nuance that language offers. And with the help of new AI tools, researchers are learning to decipher that language. They believe that soon, they will be able to speak to the whales.

It is hard to overstate what this would mean, if it happens. Please try to imagine it: for the first time, humans would be able to have a conversation with representatives of another conscious, intelligent species. One far older than ours, and embedded into a very different environment: the deep ocean.

The research behind these breakthroughs is complex (how could it not be? see sources below), but it is not complicated to understand. Sperm whales communicate using combinations of fast, rhythmic clicking sounds, grouped into identifiable clusters, which researchers call codas. A comparison to Morse code comes up often, but this is far too simplistic. The term “coda”, first used in 1977, was borrowed from music — which is appropriate, since for sperm whales, communicating with codas is all about the variations in rhythm, timing, and other aspects of their clicks.

Researchers have been studying codas for years. But recent AI-supported studies have found that there is a much larger, much more structured variability in sperm whale codas than human ears could previously discern. Researchers now use the word “alphabet” to describe the basic set of codas that whales combine to communicate — but they are not referring to letters, as in our ABCs. They are listening to phonemes, the discrete sounds that get put together into what we call words and sentences. English, for example, may have a 26-letter alphabet, but it has 44 basic phonemes that we combine to make words. By comparison, Mandarin Chinese has at least 56 phonemes, maybe more depending on how you count. (No wonder it is difficult for an English-speaker to learn Chinese!)

The equivalent basic set of sperm whale sounds has, thanks to AI-powered research, grown from an initial 21 identified codas a few years ago, to 143 codas today — which makes for many more combinatorial possibilities than either English or Chinese, at least in theory. Researchers also use other musical terms, like “rubato” (how long a coda-phrase lasts), “ornamentation” (something like little grace notes in between phrases), and “tempo” to describe and categorize these sounds. They make beautiful tables to visualize the coda system. It’s hard to grasp the details if you are not deeply immersed whale-communication research, so in trying to imagine what it might feel like to do this work, I fall back on science-fiction metaphors: you know the aliens are talking, but the way they talk is so extremely different, you end up grasping for concepts outside our usual human linguistic experience to build a bridge to understanding. Music is a fine go-between.

But the more researchers learn, the more traditional linguistic concepts also start to make their appearance. Researchers now talk about whales having “vowels”, and even “diphthongs” (bends in the pitch). The possibilities within this structured sound system for transferring information are enormous. There is no way we humans could have observed it, much less attempted to replicate it, without machine and computer help. But now, we can both observe and replicate, and researchers are already trying to build an AI-powered translator, English-to-whale. They have even started to think through how to evaluate whether the translator, when it has been completed and tested, is actually working.

AI should not get all the credit for these breakthroughs. The human research teams involved, organized by the non-profit organization Project CETI, are very dedicated and highly multi-disciplinary, comprising marine biologists, linguists, cryptographers (code-crackers), roboticists, drone pilots, and human machine learning and AI experts. AI is great at finding patterns that humans cannot see, but as you probably know from your own use of ChatGPT and other systems, you have to know how to ask the right questions, and then how to interpret the responses. You can’t just plug whale noise into ChatGPT and ask, “So what are they saying?”

So, what are the whales actually saying? Nobody knows yet, except for the fact that when whales “talk”, they do seem to include certain patterns that help identify them to others, by clan and individual. (Think of a big conference call: “Alan in Stockholm here. And I think …”) The AI systems have learned to “read” those parts of the communication that establish the whale’s identity, and they can tell which whale, from which clan, is “speaking”, with over 90% accuracy.

And of course, the more we learn about the structure and complexity of sperm whale communication, the more we can learn. Judging by the dates on recent papers, the learning curve is tipping up, quickly. Researchers are now convinced now that sperm whale vocalisations “represent one of the closest parallels to human phonology of any known animal communication system,” beyond what we can see anywhere else in nature, such as in birdsong, or even the complex songs of humpback whales.

So what’s so special about sperm whales? For starters, they have the largest brains on the planet. Researchers have been interested in them for a long time, so there was already a vast database of sperm whale recordings to start analyzing, as well as to use in training an AI system to look for patterns. There was also clear evidence, well established, that the whales were communicating with each other somehow. For example, careful observations have led researchers to the conclusion that groups of whales make group decisions, by consensus.

All those clicks we now call codas certainly seemed to be meaningful, and they were the most likely candidate for a medium by which the whales can trade information about where the giant squid are running, or confer on where to swim next. Now we have a growing stack of studies establishing that these codas are meaningful — or rather, that they are well suited to encode and transmit meaning, through multiple ways of varying and combining them. Sperm whale communications are highly structured, predictably variable, and definitely non-random.

The word “game-changer” does not begin to describe what this continuing trail of breakthroughs might lead to. If sperm whales do have a language, it could be much more interesting than ours. For one thing, “sperm-whale-ese” and its many different dialects (yes, whales have dialects) would be so much older than any human language, potentially millions of years older. And it would have developed not in our mostly flat, two-dimensional terrestrial environment, but in the vast open spaces of the 3D ocean. As Project CETI director David Gruber wrote in the New York Times (his excellent short summary is what sent me down this research rabbit hole):

”Altogether, these findings are leading us to an extraordinary conclusion: Whales may possess a communication system more intricate than our own, one that possibly predates human language by tens of millions of years.”

Which leads me to my own conclusion about how these breakthroughs could potentially change the world. Learning what whales have to say, to each other and possibly even to us, could alter everything about how we view the evolution of life on Earth. It would certainly have a major impact on how we manage global environmental issues, because the whales would eventually need (maybe even request) some kind of seat at the negotiating table, or at least enhanced representation. Indeed, the debate on the legal implications of these discoveries has already begun, at least in the academic press.

For me, however, the legal considerations are just the tip of the philosophical iceberg. True dialogue with another intelligent species would shatter, once and for all, the strange belief that humans are “alone at the top” of the evolutionary ladder, with special rights somehow accruing to us as the uncontested winners of the “World’s Top Predator” competition. At the very least, this advance in interspecies communication could, or at least should, enhance the moral imperative to act with real stewardship in our caretaking of planet Earth, and all the life on it and in it.

Let me admit a bias. Ever since I found myself, as a teenager, playing a spontaneous game of catch with a very smart dolphin, I have found it impossible not to imagine cetaceans as intelligent — potentially very intelligent. If it turns out that we can talk to them, I have a long list of questions I would like to ask, reflections I would like to share. I am trying to imagine, for example, how to explain the book “Moby Dick” to an actual sperm whale. (Assuming they were remotely interested. Mostly, we need to listen to them.)

But I will leave you now to your own speculations. What would it mean if you could talk to a whale? What would you say?

More importantly, what questions would you ask?

Sources:

“I’m a Marine Biologist. This Is How I Talk to Whales.” David Gruber, New York Times, 23 Nov 2025, https://www.nytimes.com/2025/11/23/opinion/whale-language-ai.html

“How does whale communication work?” (TED Talk), David Gruber, 2023, https://www.ted.com/talks/david_gruber_and_shane_gero_how_does_whale_communication_work/transcript

Project CETI: https://www.projectceti.org/

“Contextual and combinatorial structure in sperm whale vocalisations,” Pratyusha Sharma et al., Nature Communications, 7 May 2024, https://www.nature.com/articles/s41467-024-47221-8

“The phonology of sperm whale coda vowels,” Gašper Beguš et al., bioRxiv, 10 June 2025, https://www.biorxiv.org/content/10.1101/2025.06.09.658556v1

“Vowel- and Diphthong-Like Spectral Patterns in Sperm Whale Codas,” Gašper Beguš et al., Open Mind, 2 Nov 2025, https://direct.mit.edu/opmi/article/doi/10.1162/OPMI.a.252/133906/Vowel-and-Diphthong-Like-Spectral-Patterns-in

“On Non-interactive Evaluation of Animal Communication Translators,” Orr Paradise et al., arXiv , 17 Oct 2025, https://www.arxiv.org/abs/2510.15768

“AI Is Decoding Whales’ Communications. Could That Be a Turning Point in the Push for Their Rights?”, Katie Surma, Inside Climate News, 29 Oct 2025, https://insideclimatenews.org/news/29102025/ai-sperm-whale-communications-legal-rights/

“What if We Understood What Animals Are Saying?: The Legal Impact of AI-Assisted Studies of Animal Communication,” César Rodríguez-Garavito et al., Ecology Law Quarterly, 22 Oct 2025, https://www.ecologylawquarterly.org/print/what-if-we-understood-what-animals-are-saying-the-legal-impact-of-ai-assisted-studies-of-animal-communication/

“Consensus movements by groups of sperm whales,” Marine Mammal Science, 32(4): 1402-1415, Oct 2016, https://onlinelibrary.wiley.com/doi/epdf/10.1111/mms.12338

“How many phonemes are there in Mandarin Chinese?” https://www.quora.com/How-many-phonemes-are-there-in-Mandarin-Chinese. For a more academic-technical source, see “Mandarin Phonology”, Education University of Hong Kong, https://corpus.eduhk.hk/mandarin_pronunciation/?page_id=33