FEATURE
Ugh, AI:
Embrace or Resist Artificial Intelligence?
Artificial intelligence (AI) is rapidly transforming how we work, how we communicate, and how we analyze and conduct scientific research. Yet, while AI is a powerful tool that can advance conservation and education, it also demands significant amounts of energy and water resources, exacerbating the climate crisis. It may also reflect existing biases within the system it learns from. This article explores AI’s environmental costs, potential biases, and possible benefits, and discusses how we, in the interpretive field, need to understand AI in order to inform others of the environmental impacts and ethical implications of using it.
Tasks in the Past
In the 1990s, I worked as a naturalist at a local nature center, and part of my job entailed writing, editing, and creating a quarterly newsletter. This process was both labor intensive and time consuming. Clip art, printed blocks of text, and photos had to be cut and pasted by hand onto a paper template aligned perfectly with both horizontal and vertical margins. This was copied on a copying machine with the final draft taken to a printer company. After a couple of weeks, hundreds of paper copies were ready to be mailed. I’d coordinate volunteers to help affix sticky labels and to sort the newsletters by zip code. Finally, after weeks of preparation and printing time, I’d drive to the post office for bulk mail distribution.
Other tasks were similarly slow paced. Program preparation required reading physical scientific journals and library books; therefore, our knowledge was limited to these resources, our personal experiences, and quarterly training sessions by experts. Program registrations involved managing lots of paperwork, and communication relied on phone calls and voicemails when staff were available.
ChatGPT is ready for a prompt, which allows us to use AI at any point, but with what cost to the environment? Photo courtesy of Laura Beltran.
A Technological Divide
Of course, the computer revolution changed our workplace tasks radically. Creating a quarterly newsletter in 2025 is totally different. I email contributors to request articles. I type the newsletter or paste submitted content into a computer file, easily formatting it with spell check, font options, and more. Adding photos and graphics takes minutes. In less than a couple of hours, the newsletter is ready to send. Editing drafts, making corrections, and distributing the newsletter to a generated list of members are tasks that require minimal time.
Other tasks have also been revolutionized. Colleagues can share lesson plans or resources instantly online. Entire teams collaborate on large projects in real time, and resources are accessible from anywhere without delays.
My father, a technology nerd, lived through the computer revolution, beginning his career when computers were the size of living rooms, witnessing their evolution to the handheld devices we use today. When the internet revolution occurred, he embraced it. He saw its potential. He knew it was going to change our way of life.
But many of those in the interpretive and environmental education field did not. That included me, because when I searched on the internet for natural history topics, it yielded sparse results. I dismissed it as useless while my dad understood that overall, people in the environmental field were far behind in getting data on the internet.
Many in our field resisted the internet revolution for reasons that included: fear that computers would replace our jobs, skepticism about the accuracy of online environmental information, concerns about energy consumption, difficulty or the unwillingness to learn the new tool, and the belief that technology would discourage people from going outdoors.
Eventually, the internet revolution radically changed our workplaces and the transformation brought by computers and the internet has made many tasks seamless. So, why are we now resisting the next technological revolution: Artificial Intelligence?
AI may streamline tasks, but AI data centers use significant amounts of electricity. Photo courtesy of Laura Beltran.
Lessons from the Past
Similar to the last technological revolution, many interpreters and environmental educators are resisting the use of AI but often from a distance. These concerns are valid, especially the impact of using AI on environmental resources; however, if we don’t understand how to use the tool effectively, how can we criticize its use effectively?
Regardless of our personal feelings on the matter, AI is here, and, as my dad believes again, (and this time, I’ve decided to listen) it is already beginning to change our lives. Additionally, it’s changing at an incredibly rapid pace where even technology experts struggle to keep up. But resisting AI, similar to the past resistance to the internet, could leave us struggling to catch up in a world where our students, audiences, and business partners are already using these tools. For us, the important thing to understand is that AI is a highly energy intensive tool to streamline data processing tasks, and it isn’t intended to replace us as educators. If used responsibly, it may help us increase our impact.
However…
With each AI prompt, the equivalent of one bottle of water is used for cooling at AI data centers. Photo by Laura Beltran.
The Hard Truths
Here is one hard truth about AI: AI uses tons of energy and water. As Alex de Vries’s 2023 article in ScienceDaily notes, when one of the first models of Chat GPT was undergoing training, it was estimated to have used as much electricity as 120 average US homes in a year. AI models are stored in data centers that use significant energy to keep servers running and cool. Air conditioning and water cooling work continuously day and night to manage the intense heat generated by servers. Research shows that running AI models can use up to 17,000 times more electricity in a day than an average U.S. household, largely because of the vast data center infrastructure behind it as de Vries also discusses.
Every time we query AI, data travels from our devices to data centers using energy to power servers, routers, and other infrastructure to have quick response times. Each query also uses the amount of a bottle of water to help keep servers cool with every question we ask as discussed in Srija Chakraborty’s 2024 article in Arxiv. Each day, more and more people are using AI and, therefore, more energy, in a time when we are in a climate crisis.
AI is being integrated into the interpretive naturalist workplace, so how may we do this ethically? Photo by Laura Beltran.
Shaolei Ren, an Associate Professor of Electrical and Computer Engineering at the University of California, Riverside, discusses in his 2023 article how AI tools use a tremendous amount of water and electricity. Here, Ren offers an example of a data center’s operational water usage: on-site scope-1 water for server cooling (via cooling towers in the example) and off-site scope-2 water usage for electricity generation.
And yet, also, here are other truths about AI: it is being used to tackle large environmental challenges. As discussed in this UN Environment Programme 2022 document, AI monitors methane emissions and develops strategies for mitigation. AI monitors air pollution globally to aid in public health policies. It can evaluate the carbon footprint of a product that you buy, so you, as the consumer, may make an informed decision to buy or not to buy. Hopefully, this in turn, will pressure companies to reduce their energy consumption in making their products. Scientists in Australia are using AI to create a system to integrate all research being conducted on the Great Barrier Reef. Therefore, they will be able to analyze impacts efficiently and work more cooperatively to develop mitigation efforts from climate change as discussed in University of South Australia’s 2025 ScienceDaily article. For our own parks, AI may efficiently analyze large amounts of data to create more precise ecological management plans and ecological monitoring of species. These are just a few of the many examples where AI may aid in critical environmental research.
Is there a way to create a balance between resisting AI and fully embracing it? Photo by Laura Beltran.
What do we do?
Environmental impact solutions are not easy. Conducting research on different AI and cloud platforms can help us choose smaller AI models that use less energy and cloud providers that prioritize renewable energy. Limiting unnecessary AI tasks would consume less resources. Advocating for transparency from AI companies on their energy and water use would allow for accountability. Supporting innovation in energy-efficient AI technology and infrastructure is also significant. Finally, and this is where we are the experts, teaching others about the environmental impacts of AI can increase awareness, and hopefully pressure AI companies to be more environmentally sustainable.
And what do we do about biases and inaccuracies? Creating clear policies and ethical guidelines within our organizations can build the awareness needed to critically evaluate AI-generated content as noted in Merel Moorman’s and Tsjalling Swiersta’s 2023 article in Minds and Machines. As interpreters, we’re already trained on evaluating our programs to be inclusive, and therefore, we have the skill to do the same thing with AI content.
Do we embrace or resist this technology? This answer can only come from your own beliefs, and I’m not advocating for incorporating AI in all you do. However, I do believe that many are using AI without understanding its environmental and ethical implications. Perhaps, instead of resisting this new technology, we see our role as educating others about it. As environmental educators and interpreters, it is our role to understand what AI is, how it is being used, and, most importantly, how to educate ourselves and others about its huge environmental impact and bias tendencies.
We need to inform ourselves on how AI is not just used in scientific research, but how it is used in the workplace and what this means as far as environmental costs, possible prejudices, and potential inaccuracies. If using certain AI technologies is both energy intensive and ethically challenging, we need to create workplace policies to give staff structure and guidelines on if, when, and how to use this technology responsibly. Just like we teach about climate change, interpret the significant loss of biodiversity, and evaluate our programs to be welcoming, it’s our role to inform ourselves, our members, and our program participants about both the positive and challenging aspects of using this tool.
AI is being integrated more and more into our work places, accessed easily from our computers. Photo by Laura Beltran.
For More Information
Berthelot, Adrien, Eddy Caron, Mathilde Jay, and Laurent Lefèvre. “Estimating the Environmental Impact of Generative-AI Services using an LCA-based Methodology.” Proceedings of the 31st Conference on Life Cycle Engineering, Turin, Italy, 2024. https://inria.hal.science/hal-04346102v2
Chakraborty, Srija. “Towards a Comprehensive Assessment of AI's Environmental Impact.” arXiv, 2024. https://arxiv.org/abs/2405.14004
Chauhan, Satyam. “The Growing Energy Demand of Data Centers: Impacts of AI and Cloud Computing.” International Journal for Multidisciplinary Research 6, no. 4 (2024). https://doi.org/g82hrc
de Vries, Alex. “The Growing Energy Footprint of Artificial Intelligence.” Joule 7, no. 10 (2023): 2191–2194. https://doi.org/10.1016/j.joule.2023.09.004
Noorman, Merel, and Tsjalling Swierstra. “Democratizing AI from a Sociotechnical Perspective.” Minds and Machines 33 (2023): 563–586. https://doi.org/10.1007/s11023-023-09651-z
Venkatasubbu, Selvakumar, and Gowrisankar Krishnamoorthy. “Ethical Considerations in AI: Addressing Bias and Fairness in Machine Learning Models.” Journal of Knowledge Learning and Science Technology 1, no. 1 (2022): 131–138. https://doi.org/10.60087/jklst.vol1.n1.p138
United Nations Environment Programme. “How Artificial Intelligence is Helping Tackle Environmental Challenges.” UNEP, 07 November 07, 2022, www.unep.org/news-and-stories/story/how-artificial-intelligence-helping-tackle-environmental-challenges
University of South Australia. “Can Artificial Intelligence Save the Great Barrier Reef?” ScienceDaily, February 12, 2025, www.sciencedaily.com/releases/2025/02/250212192452.htm
