AI finds rare bird missing for more than 30 years: “This is a monumental discovery”

Though many users of AI are more familiar with its ability to respond quickly to questions, draft documents and correspondence, or edit materials, these services only scratch the surface of what the technology is capable of. Machine learning, a technical term that describes a process where an algorithm can be trained to identify certain characteristics, is another facet of AI that is not as well known as the offerings of Chat GPT and the like.

In Australia, an AI has been taught to recognize the calls of different native birds. By introducing the calls and informing the AI of the corresponding species, an algorithm was created that can hear a call and inform the user which bird it came from. The technology proved immensely successful when it was able to pick up the call of a plains-wanderer bird, which had not been spotted since 1989. According to Zoos Victoria, the plains-wanderer is critically endangered, which helps to explain why they went unspotted for so many years.

The Threatened Species program coordinator, Chris Hartnett, at the Zoo, described the spotting of the bird as “a monumental discovery,” and he hopes that it will bring a brighter future for the vulnerable species.

The rediscovery elated conservationists at Zoos Victoria and engineers at Museums Victoria Research Institute and the Queensland University of Technology who designed the AI. The AI bird finder will stay in the area where the plains-wanderer bird call was recognized until next year, giving conservationists time to research the birds in the areas where the call is picked up.

Machine learning is no simple endeavor when it comes to the natural world. Recordings, and the greater the number, the better, are needed to train the AI. In the case of the golden-legged plains-wanderer, the team had to rely on very old recordings as one had not been spotted in close to four decades.

How AI and machine learning are working to reduce illegal logging

Similar technology is being developed to hinder the trade of illegally harvested timber and wood products.

The trade of illegal wood and forest products is one of the largest black markets in the world. Illegal logging has destructive consequences for the forest ecosystem and also reduces the income of governments, which typically require taxes and duties to be paid on timber logged. The revenue generated from logging often helps fund programs and systems that ensure companies are logging in a legal area and complying with plans approved by the government to reduce environmental consequences that could result from the logging operation.

Organizations involved in wood identification programs compile samples from forests all over the world, focusing on species that are known to be of high value to create a reference database. Then, through the application of machine learning, an AI can be trained to recognize the species in real time if enough samples for that species have been fed into the algorithm. World Forest ID, an organization working on expanding sample collection, also tracks the geolocation where the sample was taken with the hope that over time, not only will the technology be able to confirm the species but also an approximate location.

One of the hallmarks of illegally transported wood is that the species on the documents do not match the shipment, and the use of machine learning to equip those tasked with verifying this information could prevent illegal timber from entering global supply chains. Agents would be equipped with a scanner that can be placed directly on the timber or a cross-section of the shipment, and if the species is in the database it relies on, it will be able to provide that information in real-time.

Get your game on! Whether you’re into NFL touchdowns, NBA buzzer-beaters, world-class soccer goals, or MLB home runs, our app has it all. Dive into live coverage, expert insights, breaking news, exclusive videos, and more – plus, stay updated on the latest in current affairs and entertainment. Download now for all-access coverage, right at your fingertips – anytime, anywhere.