The Age of Dr. Dolittle: How Machine Learning is Helping Us Understand Animals

People have been dreaming about conversing with other members of the Animal Kingdom for ages. In the past, humans have taught primates sign language and other ways of communicating with us, and for the entirety of animal domestication, we’ve taught animals simple commands that they can understand and perform. Although groundbreaking, these types of communication have always required that the animal get on our level, communicate in a way that we’ve taught them because it’s what we understand. The ability to be able to understand what their different squeaks, barks, and meows mean, however, is something else entirely.

In a superbly titled piece from December of last year[1], Ferris Jabr details how machine learning is being used to help farmers understand what their chickens are ‘saying’. People who spend a lot of time with animals tend to pick up on different nuances in their ‘speech’ patterns, but when something odd happens, they’re often left to their own wits to determine what has happened. With the new AI program, that might no longer be the case.

On the very surface, this seems like wish fulfillment, but providing communication between chickens and humans is lucrative in a very real way. Being able to quickly and accurately identify any issues with the brood helps keep egg production on track, upon which many livelihoods depend.

Although not everyone has the need or desire to understand chickens, this work certainly represents the utility of AI for our farmers. What other animals will we be able to one day, thanks to machine learning, understand?


Questions? Thoughts? Comments? Let us know!


[1] Jabr, Ferris. “Fowl Language: AI Decodes the Nuances of Chicken “Speech””. Scientific American. (2/1/18)

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.