This week’s episode focused on technologies that may help humans communicate with animals and even with babies. The desire to speak with animals has existed since ancient times, and stories about people understanding animal language appear in early texts.
A major breakthrough in this field occurred in 1967 during the Cold War, when Russian submarines operated near the United States.
The US Navy attempted to detect nearby submarines and developed hydrophone technology that successfully recorded whale sounds. The excitement was enormous, and the recording was even released on a vinyl record that became a hit. This success also contributed to the end of commercial whaling.
It turns out that whales have repeating sound patterns similar to an alphabet, meaning it can be said that they have a form of language. However, until now there has been no technology capable of processing and decoding these patterns. This may change in the future with the help of artificial intelligence. ChatGPT, for example, is a large language model designed to decode and learn languages.
Just as such models help us understand programming languages like Java and Python, they may also help decode the language of whales. There is even an assumption that animal languages are less complex than human or programming languages, which could make them easier for AI models to analyze.
Another figure whose technology may contribute to understanding animal communication is Elon Musk. Recently, he raised 6.7 billion dollars for his artificial intelligence venture xAI, including investments from parties who had previously criticized him. This move positions the company as another major player in the artificial intelligence arena alongside OpenAI, Google, and others, and it also returned Musk to the top of the global wealth rankings.
In recent weeks, multimodal AI models have been launched that can hear, see video, and understand written language. These capabilities suggest that this may be the right moment to begin trying to understand what animals are saying. The challenge lies in the need for enormous datasets, including millions of recorded sounds.
When it comes to babies, the challenge is similar. Understanding infant communication is difficult, but a startup called Nanny AI claims to be able to analyze baby cries using artificial intelligence in order to interpret their needs.