Conservation TopicsApr 12 2021
Environmental IssuesAug 29 2021
DeforestationSep 07 2021
Water PollutionOct 30 2021
The Vast Implications of Artificial Intelligence
Innovation and technology have always been the driving forces behind societal
advancement in history. Currently it is no different, as exciting new technologies have been
developed to combat the forces damaging the environment. Many innovations combine
traditional science with computer science to clean and regreen our world.
Artificial intelligence (AI) is one of the most exciting new fields in science. AI, the
simulation of human intelligence by computers, can be used to model the global climate and
make predictions on the future. Many of these models are based on the principle of machine
learning, by which computers build a model by “learning” the patterns of a large amount of
sample data (training data) and then use the model to make predications or decisions. Much of
this sample data may be provided by ordinary citizens. For example, citizens are given a data set
and asked to make specific observations about the data set, such as the number of animals in a
photo. Then, the data provided by citizens is used to train the machine learning program. This
way, anyone can contribute from the comfort of their own home. Knowing how the world will
change will better prepare us and allow us to make more focused policy. These computer models
can also be used to educate the general public on climate change, which will motivate them to
join the global push for a greener world.
Additionally, AI can be used to analyze a multitude of environmental factors, such as
carbon emissions and their impact. Based on this analysis, corporations will be discouraged from
building harmful carbon plants (Snow). AI also has a multitude of other applications, including
but not limited to reducing waste, increasing crop yields, and identifying natural disasters. It can even track illegal fishing and deforestation, thereby preserving natural environments (Mulhern).
Each of these applications would be tremendously beneficial both to the environment and society
as a whole.
Pollution and energy consumption, which have long plagued environments around the
world, could be mitigated by AI through a variety of methods. In one example, researchers used
AI to create plastic-eating enzymes that reduced the time it takes plastic to degrade by a factor of
over 400,000 (“Plastic-Eating Enzyme…”). Additionally, AI will aid in the reduction of wasted
resources by optimizing systems of transportation, agriculture, and shipping. Lastly, AI can
develop systems that promote environmental consciousness when shopping online (Hao). However, a rtificial intelligence does come at a cost. Power intensive computers are required to train and operate machine learning programs. Due to the high energy demand of AI, these computers emit large amounts of CO 2 . In fact, research shows that 300,000 kilograms of greenhouse gas emissions are emitted during the development of one big AI model (Strubell). To put that into perspective, this is the equivalent of driving 850,000 miles in an average passenger car (“Carbon Footprint Factsheet”). Some predict that by 2025, data centers will account for 10% of all electricity use. Moreover, the number of resources needed to produce ideal results from machine learning models has been exponentially growing (“The carbon impact…”).
Currently, many AI programs are used without providing real social benefit, such as advertising models.
For AI to have a net positive effect on the environment, it will have to save more energy than it
consumes. Furthermore, AI has the potential for misuse and evil when in the wrong hands.
Additionally, there is the major issue of human bias. Because machine learning models are based
principally on sample data that humans choose to input, such models could have some form of
bias that hinders them from working as optimally as possible. Therefore, it is vital that the possible benefits and consequences of AI are carefully considered before any major action is taken.
Possible solutions to the energy consumption issue include relying on renewable energy
sources, developing more efficient GPUs (Graphics Processing Units), and abstracting data. The
idea behind data abstraction is that sample data used to train machine learning models can be
trimmed down to only leave the most essential parts (Ekin). This would make AI more
sustainable while still producing desired results. In addition, a number of novel computing ideas
are being used to streamline AI, offering hope for “greener” AI.
Like most scientific advancements, AI is simply a tool used by humans. With the right
intentions and care, it has the potential to solve a multitude of societal and environmental issues.
But at the same time, brazenness and malicious intent could lead to its downfall.
“Carbon Footprint Factsheet.” Center for Sustainable Systems, 2021,
Dhar, Payal. “The Carbon Impact of Artificial Intelligence.” Nature Machine Intelligence, vol. 2,
no. 8, 12 Aug. 2020, pp. 423–425., https://doi.org/10.1038/s42256-020-0219-9.
Ekin, Annette. “AI Can Help Us Fight Climate Change. But It Has an Energy Problem, Too.”
Horizon Magazine, European Commission, 12 Sept. 2019, https://ec.europa.eu/research-
Hao, Karen. “Here Are 10 Ways AI Could Help Fight Climate Change.” MIT Technology
Review, MIT Technology Review, 2 Apr. 2020,
Labbe, Mark. “AI and Climate Change: The Mixed Impact of Machine Learning.”
SearchEnterpriseAI, TechTarget, 31 Aug. 2021,
Mulhern, Owen. “Can Ai Help Achieve Environmental Sustainability?: Earth.org – Past: Present:
Future.” Earth.org, 1 Mar. 2021, https://earth.org/data_visualization/ai-can-it-help-
“Plastic-Eating Enzyme Could Eliminate Billions of Tons of Landfill Waste.” UT News, 3 May
Snow, Jackie. “How Artificial Intelligence Can Tackle Climate Change.” National Geographic,
National Geographic, 3 May 2021,
Strubell, Emma, et al. “Energy and Policy Considerations for Deep Learning in NLP.”
Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics,