The scale of the environmental impact of Artificial Intelligence (AI) is beginning to become clearer. On the positive side, AI will no doubt be used to find ways to optimise energy use and reduce waste across various industries and be better at allocating scare resources. However, the development and deployment of AI requires substantial computational power, which in turn demands significant energy consumption and water use. The large data centres, where AI computations occur, often rely on non-renewable energy sources, contributing to carbon emissions and require substantial amounts of water keep the data centres cool. This article from The Times showing that ChatGPT uses up to 4x more water than previously thought consuming as much as 5.4 million litres of water to train Chat-GPT3 to write a 100 word email which is equivalent to the annual water consumption of 26 homes for a family of four.
There are many other AI impacts including prodcution of the hardware, such as GPUs and specialized chips, which involves resource-intensive manufacturing processes and the mining and extraction of minerals and associated electronic waste all have environmental repercussions.
At Brickendon, we advise our clients to included ESG (Environmental, Social & Governance) into any AI feasibility study to ensure that the use of AI doesn’t detract from our clients net zero and other environmental and social goals.
If you would like to discuss how we can help you with your AI strategy please contact us today via this link
Explore the latest Insights from Brickendon and ensure that your organisation is prepared.