Overview
Recent endeavors to employ artificial intelligence (AI) for deciphering animal communication, encompassing species such as bats, whales, and birds, confront a foundational impediment. A new study, spearheaded by a research team from Tel Aviv University, highlights this core issue: AI models primarily analyze the physical attributes of a sound. This analytical focus, however, does not equate to an understanding of the meaning assigned to that sound by the animal receiving it.
Research Context
In recent years, numerous attempts have been initiated with the explicit goal of utilizing artificial intelligence to interpret the communicative patterns observed in various animal species. This includes a diverse range of animals such as bats, whales, and birds. The underlying objective of these efforts has been to translate or understand what is often referred to as 'animal language' through AI methodologies.
Findings
The study, led by researchers from Tel Aviv University, identifies a fundamental problem inherent in the current approach of using AI for animal communication analysis. The core issue articulated is that AI models concentrate on the physical properties of a sound. This focus on physical attributes, while allowing for detailed acoustic analysis, does not enable the AI to grasp the meaning that the receiving animal attributes to that specific sound. The implication is that understanding the physical characteristics of a sound is distinct from comprehending its intended or received meaning within an animal's communicative context.
Research Source
Tel Aviv University