One of the many game-changing features of artificial intelligence (AI) is the potential it holds in accelerating the search for a solution to the climate crisis.
However, while purportedly holding the key to halting further damage to our planet, an AI model may simultaneously be guilty of worsening the crisis at all stages of its lifecycle. The overall effect of AI on the environment is hard to quantify, but awareness of its impact on both sides of the fight must increase if its positive potential is to be harnessed successfully.
AI – a considerable carbon footprint
One of the main areas in which AI can serve to worsen the effects of climate change is the huge carbon footprint associated with training and running AI models. This is a highly energy-intensive process that directly adds to greenhouse gas emissions, although the extent of emissions produced will vary depending on the nature of the model. For example, a large language model (LLM) supporting a chatbot such as ChatGPT will require much more energy than a smaller machine learning (ML) model that identifies patterns in research data.
However, this will only increase as the technology develops and data centres multiply: OpenAI researchers have estimated that the most advanced AI models have been requiring double the amount of computing power every three to four months since 2012 to be successfully trained.
Considering some of the largest models can produce between 300 to 500 tons of carbon dioxide during training, according to a University of Massachusetts study, this paints a worrying future for the most rapidly evolving technology on the planet.
AI models produce lots of electronic waste
Furthermore, it is not just during training and maintenance that AI can take a toll on the environment. Global levels of electronic waste are rising rapidly and the proliferation in numbers of AI models could cause this to reach crisis levels: annual e-waste production is currently on course to reach 75 million metric tons by 2030, according to the World Economic Forum.
How well do you really know your competitors?
Access the most comprehensive Company Profiles on the market, powered by GlobalData. Save hours of research. Gain competitive edge.
Thank you!
Your download email will arrive shortly
Not ready to buy yet? Download a free sample
We are confident about the unique quality of our Company Profiles. However, we want you to make the most beneficial decision for your business, so we offer a free sample that you can download by submitting the below form
By GlobalDataIf not recycled properly, e-waste can cause the contamination of water and soil supplies when dangerous chemicals such as mercury and lead are released into the environment.
So far, the AI world has received relatively little scrutiny of its environmental impact, but to stop this issue from worsening significantly, more stringent regulation surrounding the disposal of AI hardware is a necessity.
Disinformation surrounding climate change will be easily spread
The spreading of climate disinformation by AI is a side-effect that is impossible to quantify but is potentially even more damaging than increasing emissions or environmental contamination. Denial of rising global temperatures is already worryingly common thanks to social media platforms such as X (formerly Twitter) facilitating the spread of false information, and AI will undoubtedly serve to worsen this problem.
Generative AI, in particular, will enable climate science deniers to flood social media platforms with deepfake videos and pictures, aimed specifically at manipulating target audiences into doubting the authenticity of the crisis and the efforts to tackle it. In a historic election year where, national elections are expected in at least 64 countries, climate science will be a highly politicized issue around the world in 2024.
This means that we can expect to see generative AI used as a tool to stoke the flames of the debate and help convince voters to elect candidates who refuse to acknowledge the severity of the crisis.
This therefore constitutes another area in which tighter regulation of AI-produced content is required to limit manipulation of voters and the public in general. In the US and EU, which are both set for major elections in 2024, current AI regulation policies reveal a lack of adequate protection against such uses of the technology.
On the plus side It is important to mention that AI can also have a positive environmental impact. Some of the many ways its potential can be harnessed for good include helping scientists measure the rate at which icebergs are releasing meltwater into the ocean and assisting companies in a variety of sectors in decarbonizing their operations.
However; to maximize the positive potential of AI, awareness of its side effects must be raised before action can be taken to mitigate its negative impact.
Related Company Profiles
X Corp