We’ve often wondered what birds, dogs, or cats talk to each other about. But unfortunately there is no way to understand them. box Artificial intelligence (AI) helps decipher animal language and make it understandable to people? This is the challenge posed by a group of scientists.
Researchers Yossi Yovel and Oded Rechavi from the Faculty of Zoology, Wise Faculty of Life Sciences and Tel Aviv University (Israel) Sagol Faculty of Neuroscience investigated what it takes for artificial intelligence to understand communication between animals.
In a study titled ‘Artificial Intelligence and Doctor Dolittle’s Struggle’, Current Biologyset It explores the potential and limitations of Artificial Intelligence in this field.
The title of the investigation refers to a popular children’s book series that has had numerous film and animated adaptations. In this fictional story, a West African parrot named Polynesian who is bilingual because she speaks the language of her species and English, Dr. He teaches Dolittle to understand the language of birds.
Regardless of the abilities of animals, humans or artificial intelligence, researchers Understanding another genre’s communication goes far beyond just giving meaning to soundsThis is what we would do with another human language. Multiple communication factors are needed to communicate with the animal kingdom, and researchers three main obstacles.
Context
The first limit is the difficulty of distinguishing the context of animal communication. Humans have been recording and mimicking animals’ communication for decades, and researchers have been able to get responses from various animals based on this technique, without the help of artificial intelligence.
For example, a female-looking robot frog attracted real male frogs to try to mate, and a robotic fish interacted with live fish during the learning phase, influencing their movements.
Additionally, a robotic bee was able to recruit real bees to follow the ‘waggle dance’ instructions and fly to a specific location in the field. The ‘waggle dance’ is the way bees typically relay information about the location of food to each other, similar to giving instructions with hand gestures.
These examples demonstrate the potential of engineered stimuli to elicit responses and even convey information to animals in a given context. Although artificial intelligence can produce animal-like communication sounds, this is more difficult. Determine if these sounds are closer to context beyond familiar or animal-like.
After experimenting with the recordings, the AI can produce a specific bird song, but additional information is required to determine whether the bird is singing to establish territory, impress a mate, or make a vocalization that conveys something completely different.
The same thing happens with the silent communication of insects. They communicate through chemical signals. Without observing the insects’ behavior, it can be difficult to decipher whether the chemical is a mating signal, a warning of danger, or just a random chemical release.
Currently, AI needs human input and definitions to start working on animal communication, such as field recordings of bird songs. Human biases strongly influence the interpretation of these signals. Because at some point, it is necessary to assign a context to the songs of birds so that the artificial intelligence can make connections about the songs of other birds.
elicit a natural response
Animals exhibit a wide variety of behaviors and Your answers can be affected by many factors, Including current physiological states, social dynamics and environmental conditions.
Different animal species rely on a variety of sensory methods to communicate, such as sound, chemical signals or body language. Identifying specific responses related to communication, various observation techniques. Experiments cannot train animals to produce responses; because it is incompatible with natural communication and makes observations dependent on laboratory tests.
Measuring responses can also be challenging. there may not be a clear and measurable external response. People may be more prone to missing subtle responses. Artificial intelligence algorithms designed to interpret responses would be at greater risk of getting any subtle movement in response or finding correlations that don’t actually exist.
Limited context range
The third obstacle identified is Focused on limited contexts such as animal communication, alarming and courtship. This limitation reduces the ability to communicate with animals on a wide variety of topics or contexts, potentially limiting the ability to communicate between species.
At the very least, expectations may not be met if an AI translator is designed only to tell us that birds transmit the emoji equivalent of sad faces and hearts all day long.
potential benefits
Communicating with animals has potential benefits. Doing this efficiently with bees or other pollinators can benefit agriculture. The researchers also say that discerning farm animals’ moods or tendencies can help them care. Understanding noise in a protected forest can alert conservationists to an invisible threat. Research experiments could be improved significantly if mice could share certain things with humans.
Moreover: this technique, Inferences as a test for first contact with extraterrestrials. If we cannot communicate with intelligent life on this planet, what hope can we have of communicating with life on another world?
Reference work, at this link.
……….
Contact address of the environment department: [email protected]