AI and machine learning are set to revolutionize our understanding of animal communication by 2025, driven by advancements in technology and new large datasets. The Coller-Dolittle Prize, w…
AI and machine learning are set to revolutionize our understanding of animal communication by 2025, driven by advancements in technology and new large datasets. The Coller-Dolittle Prize, with cash incentives up to $500,000, reflects growing confidence in this field. Projects like Project Ceti are using machine learning to decode sperm whale clicks and humpback songs.
Historically, the lack of large, high-quality data on animal communication has been a barrier. Unlike human language models trained on vast amounts of text, animal datasets are small and fragmented. However, affordable recording devices such as AudioMoth are enabling scientists to collect 24/7 audio data from animals in the wild.
This influx of data, paired with AI algorithms like convolutional neural networks, can detect and classify animal sounds efficiently. Despite progress, challenges remain. Unlike human language, scientists do not fully understand the meaning or structure behind animal sounds. While some organizations aim to translate animal communication, most scientists view this as unrealistic, focusing instead on deciphering patterns and signals.
The field anticipates breakthroughs in data quantity and AI sophistication, moving closer to unraveling the mysteries of animal communication.