Artificial Intelligence for Conservation: Is It Worth it?

AI for Conservation graphic

The use of artificial intelligence (AI) to deal with climate change has been a hot topic for several years. In 2021, I did a blog on using AI for conservation. Since then, users have seen many benefits. But the use of AI comes with high energy and water consumption, as well as having the potential to increase electronic waste. Pitting large consumption from AI against the objectives of land and water conservation, it appears that the two could be at loggerheads. Let’s take a deeper look.

 

Artificial Intelligence

So, what is AI? There are many definitions. In an article by the European Parliament, AI is defined as “the ability of a machine to display human-like capabilities such as reasoning, learning, planning and creativity.” A commonly used definition of AI is “the ability of a computer or a robot controlled by a computer to do tasks that are usually done by humans because they require human intelligence and discernment.” In a class that I have helped teach, AI is described as the simulation of human intelligence in machines that are programmed to “think” and “learn” like humans.

 

It is important to note that all these definitions refer to thinking like humans but not replacing humans. AI is meant to support and accelerate human decision making. AI applications can’t perform tasks outside of their designed purpose.

 

Examples of AI in everyday life are recognizing speech, making decisions, solving problems, interpreting images, and creating images. Other uses of AI include reduction of energy consumption and carbon footprints, disease diagnostics in the health industry, minimization of the use of fertilizers and pesticides in farming and acting as virtual tutors in education. There are many types of AI including generative AI and predictive AI. Generative AI is a system used to create new content such as text and images. It has a solid track record given its use via ChatGPT and DALL-E, as well as functions like image interpretation. Chatbots could even be used on conservation organization websites to engage the public and answer conservation related questions. Predictive AI is used to forecast future outcomes or trends. Predictive AI has great potential; but many still question its effectiveness because of the complexity of the problems attempted to be solved, the challenges of how and where to get data for learning, and the difficulty to foresee chance events. Should you want to dig deeper into the types of AI and their uses, I recommend “AI Snake Oil – What Artificial Intelligence Can Do, What It Can’t, and How to Tell the Difference,” by Arvind Narayanan and Sayash Kapoor.

 

Conservation and Artificial Intelligence

Conservation organizations, like profit and other non-profit organizations, have various business and mission functions that can be addressed with AI. Outreach can use AI to personalize communications and more effectively solicit funds. Education can use virtual tutors and develop curriculums with AI. Human resources leverage AI to create engaging job descriptions. Chatbots like ChatGPT help people perform standard office activities like writing emails and preparing presentations. I have used Chat GPT to help write thank you notes, get lists of software packages for specific applications, write a blog, and even get a list of databases for carbon tax rates by country.

 

Mission functions are those specifically for land, water, and wildlife conservation. Today, there are many uses of AI being currently applied or being considered for conservation.

 

The Chesapeake Bay Program (CBP) Land Use/Land Cover Data Project leverages aerial imagery, LiDAR elevation data, and machine learning to produce 1-meter resolution land cover and land use maps for precision planning. Contextualizing this data supports decision-making for future actions related to the protection of the Chesapeake Bay and serves as the basis for developing the next generation of watershed and land change models. Other organizations are using similar techniques focused on tracking water loss and helping with water conservation. I personally think there are opportunities for identifying and managing wetlands.

 

AI is being used to monitor wildlife and stop poachers. Machine learning algorithms analyze images and videos to track animal populations, reducing the manual labor required for wildlife surveys. For example, Conservation AI provides advanced machine learning techniques to analyze vast quantities of data, offering invaluable insights to conservationists. Conservation AI facilitates the identification and categorization of animals, humans, and man-made objects, particularly those indicative of poaching activities, such as vehicles. On the subject of animals, researchers at MIT are using AI in an effort to translate nonhuman communication, which could ultimately help with animal protection.

 

These are concrete examples of generative AI. Predictive AI is still developing but holding great promise. The World Wildlife Foundation has collaborated with computer scientists and artificial intelligence (AI) experts to develop an advanced computer model to combat deforestation: Forest Foresight. AI predicts areas at high risk for deforestation, helping conservationists prioritize interventions and plan ranger patrols. Piloted in Borneo and Gabon, the tool claims to predict forest loss up to six months out with 80% accuracy.

 

AI Energy Consumption, Water Consumption, and E-Waste

The advancements in the use of AI for conservation and other areas have raised greater awareness about its environmental impact.

 

Training AI models, particularly large ones like GPT-4, requires substantial computational power and, consequently, significant amounts of energy. The process involves running extensive computations on vast datasets, often over prolonged periods. For instance, training a state-of-the-art natural language processing model can consume as much electricity as hundreds of households use in a year. The data centers that run the AI applications also require substantial energy. AI accounts for about 2-3% of current global electricity consumption, a figure expected to rise sharply with increased adoption, potentially reaching 8-10% of global electricity consumption by 2030. (source: weforum)

 

Water consumption associated with AI technologies is also a concern. A significant amount of heat is generated by the energy required to power AI. This heat is released into data centers for AI. Cooling down the data centers requires an enormous amount of water. If current trends continue, the global demand for water driven by AI could reach 6.6 billion cubic meters by 2027. According to Shaolei Ren, an associate professor at UC Riverside, “If you have a conversation with an AI, like ChatGPT, for 10 to 50 questions and answers, it will consume about 500 milliliters of water, or the size of a standard bottle of water.”

 

Generative AI could account for up to 5 million metric tons of e-waste by 2030, according to a new study. Leading tech companies are spending heavily building and upgrading data centers to power generative AI projects and to stock them with powerful computer chips. If the AI boom continues, the older chips and equipment could amount to 5 million metric tons of e-waste by 2030, according to the study, published October 28, 2024, in Nature Computational Science. That’s a relatively small fraction of the current global total of over 60 million metric tons of e-waste each year. However, it’s still a significant part of a growing problem.

 

These issues can be addressed. AI training, AI models, and computer hardware are being developed to be more efficient and use less energy (See this Nature article for more insight). Data centers can have more energy-efficient cooling systems and make greater use of renewable energy sources. Water conservation strategies like evaporative cooling systems are being considered. Government policies, better deployment practices, and public awareness can help with the reduction, reuse, and recycling of electronic equipment and thus address e-waste. By combining technological innovation, operational efficiency, and sustainable policies, it is possible to significantly reduce the environmental impact of AI systems. Energy efficiency will improve. Water use will be minimized. The handling of e-waste can be better managed.

 

Conservation vs. Consumption

The relationship between AI and the environment is well recognized. There is no question that AI will continue to be used in the day-to-day operations of conservation organizations. Mission specific uses of AI make conservation more effective today and in the future. There are ways to mitigate AI’s impact on the environment by minimizing energy consumption, water consumption, and e-waste. Does it make sense to use AI for Conservation? Is it worth it? I believe the benefits outweigh the adverse effects, so my answer is “Yes.”

 

Ready to expand the thinking for land and water conservation? For more than 50 years, The Conservation Foundation has been a thought leader providing education on new and exciting ways of doing conservation and applying them to everyday life. The Conservation Foundation is also a forum for folks to push the envelope and challenge people in their thinking. We are always looking for creative, hard-working people – become a member today!

 

Feel free to comment on this blog with additional ideas you have on how AI can improve land and water conservation efforts.

 

By Steve Stawarz, Oak Brook
DuPage County Advisory Council Member, The Conservation Foundation
Instructor for Everyday AI, People’s Resource Center

 

Image: generated with ChatGPT, animated with HailuoAI.

 

Like this article? Share it!

Share on Facebook
Share on Twitter
Share on Linkdin
Share on Pinterest

Winter Chloride Watchers Training Registration

Name(Required)
Training Date(Required)
Which training session would you like to attend?
This field is for validation purposes and should be left unchanged.