By: Rana Nejad

Edited By: Mustafa Ahmad

We all want to talk to our pets; ask them about their day, figure out their favorite family member, and what they do when we’re not home. To many, this idea might resemble the plot of a Pixar film. However, thanks to advancements in artificial intelligence (AI) and machine learning, this dream could soon become a reality.

Communicating with animals is not a newfound ambition. Humans have attempted to communicate with different species since the early 20th century, with indigenous populations having paved the way. 

Prior studies have examined the communication patterns of naked mole rats, revealing their remarkable communication skills. The investigations unveiled the existence of unique dialects within mole rat colonies, demonstrating a cultural transmission of these dialects across generations. Social upheavals, such as the forceful removal of a mole rat colony’s queen, were identified as catalysts for changes in these dialects, giving rise to novel linguistic patterns under the rule of a new queen.

Scientists have recently begun utilizing advanced technology to decode animal communication. Machine-learning algorithms now play a crucial role in identifying stress indicators in squealing pigs and unraveling the motives behind the vocalizations of fruit bats. Programs like DeepSqueak, capable of automatically detecting, analyzing, and categorizing the ultrasonic vocalizations of rodents, have been employed and adapted to scrutinize the communication patterns of various animals, including lemurs and whales. 

MeowTalk, an AI-powered cat translator app, has risen to fame over the past few years for claiming to maintain the ability to translate between an individual cat’s different noises to classify intent. In 2021, MeowTalk researchers reported that the software could distinguish among nine intents with 90 percent accuracy. However, developers admit that the app was better at identifying certain intents over others, often times confusing indications of happiness and pain. 

Nonprofit organizations at the forefront of animal decoding initiatives, such as the Earth Species Project and the Cetacean Translation Initiative (Project CETI), collaborate with scientists and machine learning experts to unravel communication patterns among whales and other species. The success of these projects relies on the availability of extensive data on how these species interact. Such information could empower machine learning systems to better understand and support two-way communication. 

In a conversation with AI expert Dr. Monica Lopez-Gonzalez, we explored the profound impact of AI and machine learning on reshaping our understanding of interspecies communication. Our discussion began with an overview of past efforts in interspecies communication that approached animal communication within the confines of human frameworks. These strategies largely took an anthropomorphized approach, concentrated on higher-order primates or haplorhines (including Tarsiers, Old and New World monkeys, apes, and humans). Consequently, these studies yielded limited success, as they neglected key anatomical and cognitive differences between humans and animals.

“This was the wrong way of thinking,” says Dr. Lopez-Gonzalez. Nevertheless, these studies proved exceptionally valuable by shedding light on the limitations of intelligence in higher-order primates. She notes that the progress in AI and machine learning in the contemporary era helps “remove the human construct out of the equation”, allowing for the study of animals in their natural habitats as machines document their communication patterns. In this way, combining visual data points and vocal sounds could potentially replicate a comprehensive communication system among animals soon. 

The potential benefits of decoding animal communication extend beyond casual pet conversations. Understanding the natural habitats of different species provides insights into survival strategies, contributing to the conservation of endangered species and addressing food security challenges in fishing and agriculture.

Dr. Monica Lopez-Gonzalez and other experts in the field consider these advancements in AI positively. However, like many technological advancements, the improper use of animal decoding technologies is always a possibility. Thus, serious ethical considerations around the rights of animals must be part of the picture. 

Should we be able to translate animal communication into a language comprehensible to us, it is crucial to ensure that we use this information responsibly and refrain from exploiting it to inflict harm upon a species. On the human side, the extraction of user-centric data from applications such as MeowTalk, which utilizes visual and auditory data to monitor a cat’s activities, must be monitored to uphold data privacy regulations. Before using phone apps to examine the inner thoughts of our furry friends, it is crucial to ask, “What data beyond the animal is included when gathering the data necessary to decode the communication?” 

But what about ethical concerns regarding the animals? Should we consider these animals as individual agents who own their lives? We know animals undergo emotions, endure suffering, interact, and adapt to their surroundings. Hence, there might be a necessity for a fresh animal rights framework to prompt new inquiries regarding the rights of animals in relation to computer-driven decoding programs. “Data becomes particularly crucial in this context,” remarks Dr. Lopez-Gonzalez, emphasizing the importance of hard data in shaping and supporting animal rights arguments. In this way, we may have to develop a Geneva Convention for cross-species communication that needs to be established before we acquire the ability for two-way communication with different species. 

Aza Raskin, co-founder of the Center for Humane Technology and the Earth Species Project, emphasizes that machine learning languages like ChatGPT may simulate understanding but ultimately rely on humans to ensure responsible and ethical interspecies exploration.

Importantly, I wanted to know: could AI eventually replace the work of renowned scientists such as Jane Goodall, who spent her life living and studying primates in Tanzania? What sets Goodall’s work apart from AI-powered efforts, however, is the distinctly human aspect of building trust with the animals she observed. According to Dr. Lopez-Gonzalez, recent advancements “don’t necessarily negate previous research, but simply approach it in a different way.” 

However personally fearful of an ape-led uprising, I have to admit that the prospect of interspecies communication would be a truly revolutionary undertaking. The largely unexplored potential of this field indicates that scientists have a vast and thrilling journey ahead of them, making casual conversation with our pets a promising long-term goal.