{"id":1707,"date":"2024-12-08T12:33:43","date_gmt":"2024-12-08T17:33:43","guid":{"rendered":"https:\/\/students.bowdoin.edu\/bowdoin-science-journal\/?p=1707"},"modified":"2024-12-12T09:41:58","modified_gmt":"2024-12-12T14:41:58","slug":"ai-save-or-ruin-the-environment","status":"publish","type":"post","link":"https:\/\/students.bowdoin.edu\/bowdoin-science-journal\/csci-tech\/ai-save-or-ruin-the-environment\/","title":{"rendered":"AI \u2013 save or ruin the environment?"},"content":{"rendered":"

With the fast speed that AI is currently developing, it has the potential to alleviate one of the most pressing problems\u2014climate change. AI applications, such as smart electricity grids and sustainable agriculture, are predicted to mitigate environmental issues. On the flip side, the integration of AI in this field can also be counterproductive because of the high energy demand of the systems. If AI helps us to transition to a more sustainable lifestyle, the question is, at what cost?<\/span><\/p>\n

The last decade saw exponential growth in data demand and the development of Large Language Models (LLMs)\u2013computational models such as ChatGPT, designed to generate natural language. The algorithms resulted in increased energy consumption because of the big data volumes and computational power required, as well as increased water consumption needed to refrigerate data centers with that data. This consequently leads to higher greenhouse gas emissions <\/span>(Fig.1).<\/span><\/i> For example, the training of GPT-3 on a 500 billion-word database produced around 550 tons of carbon dioxide, equivalent to flying 33 times from Australia to the UK [1]. Moreover, information and communications technology (ICT) accounts for 3.9% of global greenhouse gas emissions (surpassing global air travel) [2]. As the number of training parameters grows, so does the energy consumption. It is expected to reach over 30% of the world\u2019s total energy consumption by <\/span>2030<\/span><\/a>. These environmental concerns about AI implementation led to a new term\u2014Green AI.<\/span><\/p>\n

\"\"
Fig 1: CO2 equivalent emissions for training ML models (blue) and real-life cases (violet). In brackets, the billions of parameters are adjusted for each model [3].<\/figcaption><\/figure>\n

Green algorithms are defined in two ways: green-in and green-by AI (<\/span>Fig. 2<\/span><\/i>). Algorithms that support the use of technology to tackle environmental issues are referred to as <\/span>green-by AI. <\/span><\/i>Green-in-design algorithms (<\/span>green-in AI<\/span><\/i>), on the other hand, are those that maximize energy efficiency to reduce the environmental impact of AI.\u00a0<\/span><\/p>\n

 <\/p>\n

\"\"
Fig. 2. Overview of green-in vs. green-by algorithms.<\/figcaption><\/figure>\n

 <\/p>\n

Green-by AI <\/b>has the potential to reduce greenhouse gas emissions by enhancing efficiency across many sectors, such as agriculture, biodiversity management, transportation, smart mobility, etc.\u00a0<\/span><\/p>\n