Dashboard Export Edit

AI Trends

The environmental cost of AI-generated imagery examined

N

By Noah Newsroom

Monday, 28 October 2024 7:50AM UTC

A joint study reveals the significant energy demands of generating AI images, sparking concerns about the environmental impact as usage escalates.

In recent assessments, the environmental cost of generating images using artificial intelligence (AI) has come under scrutiny. As AI technology becomes increasingly integrated into various sectors, a joint study from Hugging Face and Carnegie Mellon University has brought to light the significant energy requirements of AI image generation. This study provides a clearer picture of the vast amount of energy consumption involved in creating AI-generated images, an aspect often overshadowed by ethical debates and technical advancements in the field.

The study estimates that producing a single AI-generated image can consume up to 0.011 kilowatt-hours of energy, roughly equating to half the battery life of a standard smartphone. However, this number is not fixed, as different variables such as the specific AI model utilised and the desired size of the image can influence the energy expenditure. Despite appearing modest on an individual scale, the cumulative effect is considerable; with estimates suggesting that around 34 million AI images are generated daily, the total energy consumption becomes significant.

This growing energy demand associated with AI-generated imagery has understandably raised concerns within environmental and political spheres. As the world grapples with climate change and seeks sustainable solutions across disciplines, the high-energy requirements of AI are seen as a critical consideration in the broader discussion on reducing carbon footprints.

The findings from Hugging Face and Carnegie Mellon University highlight the urgent need for discussions on how AI technologies can be adapted to minimise their environmental impact. While AI continues to offer impressive capabilities, the energy cost associated with these technologies presents a challenging facet that stakeholders in the tech industry will likely need to address. As the development and deployment of AI technologies continue to expand, balancing innovation with environmental responsibility may emerge as a key focus for researchers and policymakers alike.

Source: Noah Wire Services

More on this

https://www.engadget.com/researchers-quantify-the-carbon-footprint-of-generating-ai-images-173538174.html - This article corroborates the study by Hugging Face and Carnegie Mellon University on the energy consumption and carbon footprint of generating AI images, including the comparison to charging a smartphone.
https://www.inavateonthenet.net/news/article/ai-usage-puts-huge-pressure-on-energy-grids-says-report - This report supports the findings on the high energy consumption of AI models, particularly highlighting the impact on energy grids and the significant energy usage compared to task-specific software.
https://www.technologyreview.com/2023/12/01/1084189/making-an-image-with-generative-ai-uses-as-much-energy-as-charging-your-phone/ - This article details the energy consumption of generating AI images, comparing it to charging a smartphone, and discusses the broader environmental implications of using generative AI models.
https://www.slashgear.com/1696332/ai-image-generation-how-much-energy-used/ - This article provides specific estimates of the energy used for AI image generation, such as 0.011 kilowatt-hours per image, and discusses the cumulative effect of high-volume image generation.
https://www.technologyreview.com/2023/12/05/1084417/ais-carbon-footprint-is-bigger-than-you-think/ - This article highlights the carbon footprint of AI, including the energy consumption of generating images and the broader implications for the environment, especially in the context of global climate discussions.
https://www.engadget.com/researchers-quantify-the-carbon-footprint-of-generating-ai-images-173538174.html - This article explains the variability in energy consumption based on different AI models and image sizes, and the overall environmental impact of these practices.
https://www.inavateonthenet.net/news/article/ai-usage-puts-huge-pressure-on-energy-grids-says-report - This report discusses the political and environmental concerns raised by the high energy consumption of AI models, including the pressure on data centers and energy grids.
https://www.technologyreview.com/2023/12/01/1084189/making-an-image-with-generative-ai-uses-as-much-energy-as-charging-your-phone/ - This article emphasizes the need for transparency and informed decision-making regarding the environmental impact of AI models, highlighting the difference between training and usage emissions.
https://www.slashgear.com/1696332/ai-image-generation-how-much-energy-used/ - This article mentions the political actions and legislative proposals aimed at addressing the environmental footprint of AI, such as reporting requirements and environmental impact ratings.
https://www.technologyreview.com/2023/12/05/1084417/ais-carbon-footprint-is-bigger-than-you-think/ - This article discusses the importance of understanding the carbon footprint of AI in different regions based on the cleanliness of the power grid and the potential for reducing emissions by using renewable energy sources.
https://www.technologyreview.com/2023/12/05/1084417/ais-carbon-footprint-is-bigger-than-you-think/ - This article highlights the broader context of climate change and the need for sustainable solutions, emphasizing the role of AI in this discussion and the need for balancing innovation with environmental responsibility.
https://www.slashgear.com/1696332/ai-image-generation-how-much-energy-used/ - Please view link - unable to able to access data