Researchers are working on an AI model that would allow humans and animals to speak to one another for the first time.
Artificial intelligence is becoming more advanced as investments in the technology increase and more data is fed to the various models. With recent advancements in AI capabilities, a language model could enable interspecies communication. Researchers have been racing to solve the puzzle of animal-to-human translation, as the one who figures it out could win half a million dollars for the Coller-Dolittle Prize.
The Coller-Dolittle Prize
— Coller Dolittle Prize (@CollerDolittle) November 26, 2024
The recently developed award aims to encourage scientists to figure out a way for humans to talk to animals. It is set to be an annual award, with the first and subsequent awards until the code is cracked to be given for progress made towards achieving the ultimate goal of interspecies communication.
The award offers a prize of either $500,000 or a $10 million equity investment to the researcher or team who completes the mission. The reward for recognizing progress made towards the end goal is set at $100,000. All of the prizes can be split among two people upon request. To meet the criteria, scientists working on the project must submit a two-minute video to be shared publically on social media, along with an up to five-page paper detailing their research formatted to font size 11 and 1.5 line spacing.
The race for AI enabling animal-to-human translation
Since the dawn of artificial intelligence, different groups of researchers have been working tirelessly on a way to enable interspecies communication using large language models (LLMs). Two examples of extensive work done to achieve this goal are Project Ceti and the non-profit Earth Species Project.
Project Ceti has been using AI to attempt to translate the click communication used by sperm whales and the song-like language of humpbacks. The collective of researchers has been working on the project for about a decade and has made some progress toward understanding whale communication.
The project, along with others attempting to achieve animal-to-human communication, requires more data. Project Ceti has about 8,000 whale vocalization sound bites, which is a small sample size compared to the training data for models like ChatGPT.
Meanwhile, the Earth Species Project is trying to achieve interspecies translation on a more general and non-aquatic level. The team, which is “dedicated to using artificial intelligence to decode non-human communication,” is aiming to solve the general problems with animal-to-human communication. They have a large collection of different animal sounds and are working on deciphering crow language.
While progress is being made to achieve interspecies communication, many difficulties remain. The potentially revolutionary concept will likely take many years for even AI to master.
“AI and deep-learning tools are not magic,” said Kevin Coffey, a neuroscientist who helped create DeepSqueak, an AI model attempting to decipher ultrasonic animal communication. “They are not going to suddenly translate all animal sounds into English. The hard work is being done by biologists who need to observe animals in a multitude of situations and connect the calls to behaviors, emotions, etc.”
See all the latest news from Greece and the world at Greekreporter.com. Contact our newsroom to report an update or send your story, photos and videos. Follow GR on Google News and subscribe here to our daily email!
