The text on this page was automatically translated and hence may differ from the original. No rights can be derived from this translation.
Why is the subject important?
What makes these studies truly fascinating is how topical and widely discussed the subject is. The impact of AI on energy consumption and emissions is receiving a lot of attention - opinion pieces and articles are published almost daily. Consequently, it was a great challenge to organise all this information: which studies are truly still relevant in this rapidly changing context?
In opinion pieces and articles, concerns are regularly raised about the sustainability of AI, but there are also many misconceptions. The current studies are therefore important to examine existing views on AI and data centres. This helps to gain a better understanding of the actual environmental impact of AI, allowing for better decisions on how to make AI more sustainable.
What did you find interesting about the subject?
Emma: I found this research personally very interesting because we looked not only at the current situation, but also ahead: how will AI develop in the coming years, what does that mean for energy consumption and emissions, and what technological, economic or policy factors will play a role? With this forward-looking perspective, the research could truly contribute to the broader debate on sustainability and digitalisation.
Iris: In addition to investigating the impact of AI and data centres on sustainability, it is also important to zoom out to consider the impact of the digital sector as a whole. The research I contributed to focused on a broader perspective: mapping the emissions, water and raw materials consumption of the entire digital sector in the Netherlands. To investigate this thoroughly, it is crucial to first determine what exactly falls within the digital sector - which components are part of it, and in which processes do most emissions occur? Answering these questions provides an overview of the sector as a whole, allowing for comparison of the impact of data centres with other parts of the digital sector, such as network usage and digital equipment.
How did you approach the studies?
Emma: For this research, Dialogic brought together knowledge and expertise from various sources. Six colleagues participated in the project, and we collaborated with experts from Utrecht University and a data centre specialist. Combining everyone's in-depth analyses into a coherent report was a challenge, but it certainly contributed to the quality of the research. Dialogic demonstrates its ability to explore complex issues thoroughly and from multiple perspectives.
Iris: We systematically analysed the digital sector in the Netherlands in the research, using 'impact pathways'. In essence, this involved determining the impact - in terms of emissions, raw materials consumption, and/or water usage - of each component and sub-component of the sector. Delving deep into this analysis is very interesting. Through this systematic approach, we could identify which parts of the digital sector are most relevant to monitor.
How did the research make you think?
Emma: Both studies revealed that monitoring sustainability in both AI and the entire digital sector is still in its early stages. Many more steps need to be taken to accurately assess the impact. For instance, when it comes to data centres, information on energy consumption is to some extent available through EED reports, but there is still a significant challenge in documenting the capacity of data centres used for AI. Furthermore, information regarding the energy mix of data centres is also limited.
Iris: The digital sector generates emissions, but can also contribute to sustainability. AI can be used to make processes more efficient, thereby reducing emissions. When monitoring the digital sector, it would be great if we could assess both the emissions and the savings on emissions.
Want to learn more about the research? Check out the studies on 'De impact van AI op duurzaamheid en het monitoren daarvan' and 'Data over duurzame digitalisering'.