In the 1970s, Thomas Nagel famously asked “what is it like to be a bat?” Today, large language models clamor to give an answer. Researchers optimistically argue that AI will allow humans to speak to whales, monkeys, even bats within a handful of years.
However, interspecies communications experts have pointed out major flaws in such bold claims. To explore the historical, philosophical, and empirical parameters of communicating between species, SFI co-organized the working group “Interspecies: Decoding, Translation, and Interpretation” in collaboration with the Interspecies Internet. Held at SFI May 4–6, 2026, the event expanded on a previous meeting about animals in translation.
Participants discussed what it means to “decode” communication, extracting meaning from deep databases of complex, imperfect data. They reviewed case studies: octopi experiencing something akin to nightmares, researchers playing sounds to divert a whale from a danger zone, apes communicating through gesture, wolves holding complex social relationships.
“The field has focused on figuring out nuts and bolts: what does a chimpanzee or a killer whale or a nightingale have in their communicative toolkit to find a mate, food, somewhere to sleep? But that’s probably the least interesting part of what we humans do with our communication, and it might be the least interesting part of what these other animals do,” says working-group participant Cat Hobaiter, a professor of origins of mind at the University of St Andrews who researches great-ape communication.
Participants considered how behaviors like song or social interactions illuminate possible creativity or cultural differences in other species. Decoding communication could help solve the problem of other species’ minds — revealing a consciousness related to, but distinct from, our own.
In contrast, participants also emphasized how important it is not to overstate the progress AI models and other attempts at interspecies communication have made. Anthropomorphism — for instance, assuming most animals primarily “speak” a “language” through vocal sounds — is an ever-present risk.
“There’s no simple system to crack the code of animal communication, and that’s why we need multiple meetings to approach solving the problem,” says working group participant Irene Pepperberg, a research professor in Psychological and Brain Sciences at Boston University who has demonstrated the advanced communicative and cognitive abilities of grey parrots, starting with one named Alex in the 1970s.
“Looking at the original interspecies-communication researchers, we’re an endangered species ourselves. And nasty disagreements destroyed the field initially, so it was great at SFI to see everyone working on a path forward,” says Pepperberg.
Working-group organizers are planning a sequel to a particularly contentious 1980 New York interspecies-communication meeting, hoping to symbolically mark a new era of collaboration in the field.
“The challenge of translating from one species to another vastly exceeds the already extraordinarily difficult problem of translating from one human language to another. This becomes doubly challenging when one considers that the content of communication is likely to be a form of social coordination and not a utilitarian description of the world,” says SFI President David Krakauer, a co-organizer of the meeting. “The fascination of this problem is considering very new theories that go beyond the rather simple fitness-enhancing picture presented by the standard Darwinian framework. The new data make this a very exciting moment for new discoveries.”
This working group received support from the Miller Omega Program at SFI and the Interspecies Internet.