The environmental impacts of AI need to become more visible
© fstopimages/Malte Müller
Artificial intelligence (AI) has been on everyone's lips since the first version of ChatGPT was released at the end of 2022. Intelligence software applications have already been in use for many years to, for example, compile favourite playlists, weed out spam emails, find the best route to take, translate texts and get product recommendations based on previous internet purchases. But what about the carbon footprint of AI? Jens Gröger, Senior Researcher in the Products & Material Flows Division, explains the environmental advantages and disadvantages of AI in Oeko-Institut's latest podcast.
More efficient processes vs. increased energy demand
In contrast to traditional computing, machine learning is based on very large amounts of data and parallel computing processes. This is accompanied by increased computing power. Currently, approx. 1.5 percent of Germany's electricity demand is used for data centres alone. This demand will continue to rise in the future. This is because computer applications are currently being equipped with more and more AI functions. For example, a query via ChatGPT consumes three times as much electricity as a traditional search query. If AI functions also find their way into normal office applications, such as text and image editing programs, their need for electricity will increase considerably. The environmental impacts occur both during the training and operation of AI systems. The training of ChatGPT in Version 3 alone is estimated to have caused 500 tonnes of CO2, with a single request accounting for around 4.5 grams of CO2.
At the same time, AI has the potential to optimise technical processes – such as the production, maintenance, use and ultimately waste sorting and reuse of products – to help save energy and resources and to promote the circular economy. AI can also help optimise the use of wind and solar power in the energy sector. But do the positive effects outweigh the disadvantages? According to Gröger, these questions are still unanswered in many cases and require both further research and legal regulation.
Measuring the environmental impacts of AI
To measure the carbon footprint of AI, experts distinguish between three levels:
- Direct effects that can be directly attributed to digital technology. These comprise the production and use of devices, data lines and data centres.
- Indirect effects associated with the use of digital applications or AI. In terms of online shopping, for example, these are packaging and delivery; in terms of optimising production processes, these effects include a reduced energy demand.
- Systemic effects that affect society as a whole, such as changes in mobility behaviour due to car-sharing services or changes in the world of work. These also include rebound effects, i.e. savings in one place that are accompanied by increased consumption elsewhere.
The environmental impacts of the direct effects are the ones that can be best calculated. Indirect effects can be estimated based on use cases. Systemic effects have been difficult to quantify to date. Jens Gröger is in favour of a life cycle assessment approach: “In a life cycle assessment, we examine the entire life cycle of a product, from raw material extraction and production to transport, use and disposal. This methodology can also be applied to digital applications such as software and AI.”
Transparency as a basis
With the knowledge of the environmental impact of digital applications, energy consumption can then be reduced in a second step.
When it comes to digital technology and AI, we can't just let technical development run its course. It can go in the wrong direction. A technology impact assessment and regulation based on this are essential. Undesirable developments should be recognised at an early stage before they become uncontrollable.
The researcher argues in favour of providing environmental product information with every digital service, e.g. in the form of a small data package with information on energy and resource consumption as well as greenhouse gas emissions. Users and, above all, companies subject to reporting requirements can then track and evaluate their carbon footprint and other environmental impacts and take appropriate measures to improve their balance sheet.
Knowledge instead of everyday advice
Oeko-Institut's podcast “Wende bitte!” (“All change please!”) is aimed at listeners from politics, science, the media, NGOs and the general public – anyone with an interest in political and environmental issues. The podcast is hosted by Mandy Schoßig, Head of Public Affairs & Communication, and Hannah Oldenburg, Digital Communications & Social Media Manager at Oeko-Institut. For around an hour – enough time for the “long haul of environmental podcasts” – they talk to one of Oeko-Institut’s experts about upcoming transformations towards sustainability. Current political and social issues are addressed in the special episodes.
The “Wenden bitte!” (“All change please!”) podcast: Episodes of Season 4
All seasons and episodes of the podcast are available at www.oeko.de/podcast
The podcast is also available on all common podcast portals such as Apple Podcasts and Spotify.
