While artificial intelligence’s use of natural resources has raised concerns, it has also contributed to the development of numerous environmental solutions. So, is artificial intelligence good for the environment or bad? Artificial intelligence has been touted a lot as a good way to fight climate change.
However, tech companies ought to take measures to ensure that artificial intelligence infrastructure does not contribute to the destruction of the environment.
Modern society is rapidly adopting artificial intelligence (AI). It has rapidly integrated itself into search engines, social media, and technologies like GPS navigation in addition to being restricted to chatbot operations. However, as AI usage has increased, environmental costs associated with AI operations have raised concerns. The artificial intelligence industry is getting closer to an energy crisis, according to OpenAI CEO Sam Altman.
Why AI is a resource-hungry beast Each new large language model (LLM) has had its parameters vastly improved, allowing for greater flexibility and accuracy. OpenAI’s GPT, for instance, had 100 million parameters in 2018, its next version had 1.5 billion, and its third version had 175 billion. There were 540 billion parameters in Google’s PaLM LLM. As a result, these LLMs operate through enormous data centers and require several thousand high-performance chips.
These data centers produce a lot of heat, and the infrastructure for cooling them is powered by fossil fuels. It is anticipated that AI data centers account for between 2 and 4% of global emissions of greenhouse gases. Also, data centers use about 1% of the world’s electricity, which is expected to at least double in the next ten years. Due to the large number of users, the deployment phase of AI models consumes significantly more power than the training phase. When compared to a similar request in Google’s search engine, a request to ChatGPT can consume 100 times more energy.
To cool the processors, AI systems need a lot of fresh water in addition to energy. According to a lawsuit filed by Iowans against OpenAI’s GPT-4 facility, the cluster used up 6% of the district’s fresh water supply. This occurs at a time when there is a severe water shortage in many regions of the world. E-waste is also anticipated to be produced in significant quantities by the industry.
Legislators are also paying attention to these issues, which are becoming increasingly obvious. The Artificial Intelligence Environmental Impacts Act of 2024 was introduced by US senators at the beginning of February. It requires national assessment organizations to develop frameworks for evaluating the environmental impact of AI.
The application of AI may also result in additional environmental issues. For instance, AI technology is utilized by numerous prominent fossil fuel companies to boost production. Additionally, the use of AI in targeted advertising increases consumption beyond what may be necessary for a sustainable society. Can AI be viewed as a blessing in light of the vast gray area surrounding its use and the resource scarcity that is affecting numerous facets of modern society?
Anuj Mudaliar Anuj Mudaliar Assistant Editor – Tech, SWZD April 22, 2024 Technology and sustainability coming together Artificial intelligence has been touted numerous times as a promising strategy for combating climate change.
However, tech companies ought to take measures to ensure that artificial intelligence infrastructure does not contribute to the destruction of the environment.
Modern society is rapidly adopting artificial intelligence (AI). It has rapidly integrated itself into search engines, social media, and technologies like GPS navigation in addition to being restricted to chatbot operations. However, as AI usage has increased, environmental costs associated with AI operations have raised concerns. The artificial intelligence industry is getting closer to an energy crisis, according to OpenAI CEO Sam Altman.
Why AI is a resource-hungry beast Each new large language model (LLM) has had its parameters vastly improved, allowing for greater flexibility and accuracy. OpenAI’s GPT, for instance, had 100 million parameters in 2018, its next version had 1.5 billion, and its third version had 175 billion. There were 540 billion parameters in Google’s PaLM LLM. As a result, these LLMs operate through enormous data centers and require several thousand high-performance chips.
These data centers produce a lot of heat, and the infrastructure for cooling them is powered by fossil fuels. It is anticipated that AI data centers account for between 2 and 4% of global emissions of greenhouse gases. Also, data centers use about 1% of the world’s electricity, which is expected to at least double in the next ten years. Due to the large number of users, the deployment phase of AI models consumes significantly more power than the training phase. When compared to a similar request in Google’s search engine, a request to ChatGPT can consume 100 times more energy.
To cool the processors, AI systems need a lot of fresh water in addition to energy. According to a lawsuit filed by Iowans against OpenAI’s GPT-4 facility, the cluster used up 6% of the district’s fresh water supply. This occurs at a time when there is a severe water shortage in many regions of the world. E-waste is also anticipated to be produced in significant quantities by the industry.
Legislators are also paying attention to these issues, which are becoming increasingly obvious. The Artificial Intelligence Environmental Impacts Act of 2024 was introduced by US senators at the beginning of February. It requires national assessment organizations to develop frameworks for evaluating the environmental impact of AI.
The application of AI may also result in additional environmental issues. For instance, AI technology is utilized by numerous prominent fossil fuel companies to boost production. Additionally, the use of AI in targeted advertising increases consumption beyond what may be necessary for a sustainable society. Can AI be viewed as a blessing in light of the vast gray area surrounding its use and the resource scarcity that is affecting numerous facets of modern society?
Learn More: Where AI Can Make a Positive Environmental Change Despite growing concerns about AI’s impact on the environment, the technology has tremendous potential for combating climate change and controlling other polluting sources. Since approximately 4 billion people live in climate change-prone regions, the World Health Organization (WHO) recommends that every resource be utilized to avert the crisis.
For instance, AI can have a significant impact on projects to replant trees, particularly when combined with other technologies like drones. Optimized seed bombing operations are something they can do, especially in hard-to-reach places. Additionally, AI can track deforestation and natural carbon stores using satellite images.
In addition, iceberg movement and ice cap melting can be tracked more quickly and accurately using AI and satellite imagery. Similar methods can assist heavy industries in tracking and reducing their carbon emissions by as much as 30 percent. AI is increasingly being used by governments to monitor air quality and methane emissions.
In order to incorporate machine learning into their projects, non-profit organizations are collaborating with Google. As a means of mitigating the effects of climate change, the United Nations also uses artificial intelligence to predict weather patterns, implement reforestation projects, and manage waste for vulnerable communities in Africa.
The operation of renewable energy infrastructure and smart grids is becoming increasingly dependent on AI. When AI equipment is analyzing real-time data from sensors and meters, it can function at its maximum potential. The new technology can also significantly help with sustainable agriculture and water management, reducing waste and increasing the quantity and quality of produce.
AI can be used to better predict weather volatility in light of the severity of climate change. Tracking natural disasters like unseasonal rain, hail, and storms requires this. In terms of optimized recycling operations and ocean cleanup projects, AI can also significantly enhance waste management.
Guidelines for the AI Industry’s Sustainability AI has a lot of potential to solve the problem of global warming. However, its current environmental impact should not be overlooked. Companies in the technology industry ought to take the initiative to put their resources to use in resolving issues related to artificial intelligence. Here are a few ways AI can be made to be better for the environment.
Government intervention: The regulatory bodies that push for the establishment of laws regarding sustainability and carbon emissions as well as financial incentives for businesses to employ eco-friendly practices are a crucial component of the puzzle.
Transparency: Developers and businesses can get information about how much energy AI uses from AI companies. This may result in research on various AI systems and associated energy-saving solutions.
HVAC systems: Despite its widespread use, air conditioning is not long-term viable. The industry’s efficiency can be significantly enhanced by investing in alternative research.
Energy from the sun: Another way to reduce greenhouse gas emissions and promote sustainability in the AI sector is to make investments in the construction of renewable energy infrastructure.
Choosing and optimizing models: Instead of developing new models with ever-increasing parameters, AI companies must optimize existing models, which could increase resource efficiency. Additionally, energy efficiency can be improved by using smaller models for simpler tasks rather than LLMs for everything.
Algorithm and hardware use: AI companies need to set algorithms that do not need to produce results that are more accurate than necessary and optimize their hardware to operate at lower speeds. Additionally, sustainable AI use cases can be driven by locating a particular algorithm that is effective in a particular task.
The value of artificial intelligence solutions in combating climate change may emerge as they improve environmental problem solving. In the coming months and years, combining climate policy and AI policy will play a significant role. To get the most out of this new but volatile technology, every organization’s value system must include establishing ethical guidelines for the use of AI to avoid harming the climate.
