How AI has reshaped the wildlife conservation movement

By Sashrika Pandey

Photo Credit: Tobin Rogers

     When artificial intelligence (AI) is used to combat large scale problems, models are first exposed to small sets of training data before being fed more substantial sets of data to analyze. After years of considering social and ecological problems from multiple angles, researchers have compiled a plethora of information that can be evaluated, yielding results that can then be used by human researchers to identify areas where immediate aid is effective. Wildlife conservation, for instance, is an issue that continues to plague our planet, but once researchers turn their attention to the overlap between AI and ecology, new opportunities emerge.

     In an interview with Forbes, Shahrzad Gholami discusses her background in wildlife conservation as part of Teamcore, a research group at the University of South California. When asked about her perspective on applying AI to other fields, Gholami states, “We need more interdisciplinary research by joining forces with other domain experts… Lots of AI researchers want to do impactful work, but they don’t know how to find it… And people with real-world problems don’t realize that AI can help them.” Data analysis and ecology are often separated into distinct fields, but there are numerous convergences where using machines can help ecological researchers work efficiently. A partnership between human researchers and computational power can lead to a much faster response to constantly changing wildlife populations.

     Take, for instance, the work of research coordinator Jenna Stacy-Dawes at the San Diego Zoo’s Institute for Conservation Research. By using the software Wildbook, researchers have been able to determine the populations of giraffes by training a model to examine photos. This process has rapidly decreased the time necessary to comprehend the changes in giraffe populations for researchers, which is essential when fluctuations in the population size can have drastic effects later down the line. National Geographic explains that aerial surveillance of giraffe populations isn’t feasible due to the high cost and extensive amount of time spent. The use of AI algorithms, therefore, could serve as a potential solution for researchers that work on time-sensitive and data-heavy projects.

     While AI may seem like an infallible solution, data analysis with AI presents its own challenges. In an article discussing the use of AI in a study where researchers were confronted with ample data but limited funds, Nature adds that using AI doesn’t mean that the analysis will be error-free; rather, training a model with several sample sets and then testing it by generalizing it to other populations can give insight into its accuracy. Additionally, software developer Peter Ersts at the American Museum of Natural History’s Center for Biodiversity advises against wholly relying on AI for research practices and emphasizes cooperation between humans and machines.

     The loss of wildlife is a serious problem that is associated with numerous ecological issues on our planet. Mitigating the extent of this problem is a task that researchers and machines alike can work towards. The use of data analysis in this sector does raise questions about how AI can be used in other fields to combat areas where progress appears stagnant. Consequently, the widespread use of data analytics in supplementing the work of researchers is sure to spur interdisciplinary cooperation.

Forbes - “How AI Can Stop Wildlife Poaching”

Nature - “AI empowers conservation biology”

National Geographic - “How artificial intelligence is changing wildlife research”