Jeran Cloete from Stellenbosch University and Dian Spear, Jessica da Silva from the National Biodiversity Institute of South Africa, Lavhelesani Dembe Simba from the University of Fort Hare and Peter J Carrick from the University of Cape Town discuss the use of AI in conservation efforts.
Conservationists analyze many volumes environmental data in their work. For example, they may need to process decades of weather data or the movements of millions of insects. Until now, those scientists and decision makers have had to manually find and organize information, then use statistical tools that often oversimplify the source information.
Artificial intelligence (AI) tools now promise to help with all that. But can they fulfill the promise?
They are far from perfect. It was shown that they can make information with confidence and increase the hidden bias in their training data. And different AI tools have different uses, strengths and weaknesses. They need to be chosen carefully.
AI has featured among the top 10 emerging issues in biodiversity conservation in South Africa in recent times. horizon scan which we did. As part of a group of 14 experts in biodiversity conservation, we drew on discussions within our various professional networks, literature and news trends to identify issues that are likely to emerge and intensify over the next five to 10 years.
The issues fell into three main groups: technological disruption, regulatory complexity and infrastructure impact.
Among them, AI has been seen as both an opportunity and a threat to biodiversity conservation in the future.
The possibilities of AI
Our scan brought to light the power and pitfalls of AI for the type of work we do.
Another potential use of AI is in tracking. Tracking animals and insects at scale is important for conservation decisions. Birds and whales migrate across the planet every year, and insect numbers change seasonally by the billions. Image recognition AI can process camera capture data to help flood similar databases Wildlife Insights and provide information about animal behavior to aid prediction the effects of global processes such as climate change and the development of biodiversity industries.
Mass surveillance also records people sharing those spaces with animals. This surveillance can be used to detect illegal wildlife harvesting (poaching) or to prevent it human-animal conflict.
Land use is another area of conservation where AI offers opportunities. Using economic data and geographic information, custom AI models can be trained predicting deforestationwhich allows preventive action, or choose the world with maximum savings at the best price.
The complexity of ecosystems must be summarized and summarized in maps and categories to inform landscape-level decisions. Using AI increases the amount of data that can be summarized.
Chatbot is one type of AI tool that can disseminate information from large amounts of text. For example, they can be used to monitor product listings as well find wildlife trade online when it happened. They can read hundreds of scientific publications to help determine what species are in them danger of extinction. They can use many different sources to conduct an environmental impact assessment; basis for global development decisions, providing an attractive shortcut around the time-consuming reporting process.
But we’ve also seen downsides and risks.
Accidents
Local communities living off the land may see mass surveillance as an intrusion. The alienation of local communities in this way may cause them to oppose environmental management as well destruction technology in the field to protect their privacy.
Another challenge is that the technology itself has limitations. Using AI to track animals means specially training image and audio recognition systems to work with each ecosystem and piece of hardware. An AI model is only as good as the effort put into teaching it. For example, training a model on recordings from a city may cause it to ‘hear’ pigeons everywhere, producing a reliable but incomplete list of birds from natural spatial data.
Another concern with AI is that replacing human involvement could lead to job losses. If it is used to identify animals, it can affect continuity a decline in taxonomy knowledge which is particularly difficult in biodiversity-rich, low-income African countries. That information is essential for developing and optimizing AI systems.
We also found reasons for concern in land use applications.
The danger is that using AI tools in mapping may cut the map from realism down to substituting human judgment and selecting data sources compatible with AI methods. A skilled ecologist studying an ecosystem will see unanticipated surprises during the planning phase. For example, talking to local people may reveal an increase in organized farming or wildlife harvesting. An AI system can miss this important context because it can only read digitized information.
The AI cannot detect animals that run away from the cameras or identify animals that were not expected to occur in the area (images they were not trained on). And it cannot talk to people to find out their intentions or to reveal natural intelligence from their ancestors.
Chatbots also need to be used with caution. They can generate or embed fictional information. Even when they use real information, they often show a bias in their training data, preferring research and ideas from well-represented institutions in the global north, where publications have historically been dominated. men in universities earn more.
Inappropriate use of chatbot recommendations can lead to poor environmental decisions. For example, it may suggest planting trees without considering the diversity of ecosystems such as Africa savannah grasslands.
Using chatbots as a shortcut to summarize information and inform conservation decisions in Africa will reinforce colonial systems and marginalize communities and traditional knowledge.
Careful use of AI
Strict regulation of the use of AI in the natural sciences is therefore an ethical and legal obligation. The industry needs clear safeguards, standards and oversight mechanisms to prevent erroneous or inappropriate AI results from influencing decisions. It requires:
- authentication protocols to obtain the information made
- restrictions to prevent chatbots from exceeding human information and opinions
- mandatory disclosure of AI instant histories
- standards for describing the training dataset to select appropriate models.
The AI explosion presents a powerful opportunity for conservation if we use the right tools with care. When we replace human judgment with untested automation, we risk becoming their tools and the very systems we have built.
By Jeran Cloete, Dian Spear, Jessica da Silva, Lavhelesani Dembe Simba and Peter J Carrick
Jeran Cloete is a PhD student in conservation and entomology at University of Stellenbosch. Dian Spear is a senior research scientist at University of Stellenbosch. Jessica da Silva is the principal scientist at the school South African National Center for Biodiversity. Lavhelesani Dembe Simba is a lecture of entomology at the University of Fort Hare. Peter J Carrick is a distinguished researcher at the Institute for Communities and Wildlife in Africa, of University of Cape Town.
Don’t miss out on the information you need to succeed. Sign up for Daily BriefSilicon Republic’s digest of must-know sci-tech news.
