Artificial Intelligence (AI) is undoubtedly one of the most discussed topics in the tech world today, currently experiencing what is known as the “peak of inflated expectations” in the Gartner Hype Cycle. This phase often leads to future disappointment due to unrealistic expectations set on the technology. The potential risks and benefits of AI have been the primary focus, often overshadowing its existing issues such as bias and environmental impact.

One of the key concerns surrounding AI is its considerable carbon footprint, particularly during the inference phase when used on a large scale. The environmental impact of AI, especially during the training phase, is significant. A study conducted in 2019 estimated that training just one large language model could produce the same amount of CO2 as 125 round-trip flights between New York and Beijing. This fact alone showcases the substantial environmental concern that comes with the widespread use of AI.

While the current hype around AI is significant, some experts are hopeful that AI, particularly generative AI, will transition into the “trough of disillusionment” phase of the Gartner Hype Cycle. This phase often leads to a decrease in public interest as experiments and implementations fail to deliver. This could potentially mitigate some of its negative impacts, including its environmental footprint.

The hype surrounding AI technology has arguably diverted attention from its downsides. Not only are we looking at the environmental impact, but we also need to address issues of bias, lack of transparency, and accountability.

The current hype could, however, lead to a more realistic understanding of the technology as disillusionment sets in. This could pave the way for more practical applications of AI, reducing the risk of becoming overly dependent on a technology that is still evolving and has a continuously growing environmental footprint.

The dream of “AI everywhere” could potentially lead to a world highly dependent on this technology. However, it’s crucial that we remain cognizant of the possible negative impacts, including a significant environmental footprint. As we continue to explore AI’s potential, we must also tackle its drawbacks head-on, ensuring that the technology evolves responsibly and sustainably.

Science4Data is committed to cut through greenwashing and measure real impact. Join the journey to a sustainable future. Your actions matter.