The article study: How artificial intelligence can become more sustainable first appeared at the online magazine Basic Thinking. You can start the day well every morning via our newsletter update.
Artificial intelligence is an enormous energy guesser. But AI not only needs a lot of electricity, but also water – especially for the cooling of data centers. A new meta study gives recommendations on how artificial intelligence can become more sustainable.
Artificial intelligence has taken up enhancement enormously in recent years. Because: More and more people integrate AI applications into their everyday life-in from chatbots, translation services, when shopping or with smart home.
But such everyday helpers also have their price. The more people use AI supported services, the higher the energy and water consumption of the systems. For 2024 alone, the additional electricity consumption of data centers for the operation of AI-optimized servers is estimated at around 261 terawatt hours. This number should be Forecasts rise to 500 terawatt hours.
With the increased electricity requirement, however, the amount of water that is necessary to cool the data centers is also increasing. Because AI applications consume per kilowatt hour Between 1.8 and twelve liters of water – depending on the location and cooling technology.
Researchers from the Society for Computer Science have taken on this problem. In one Meta study have analyzed the ecological effects of AI and developed concrete recommendations for action in order to make artificial intelligence more sustainable.
How can artificial intelligence become more sustainable?
For their study, the researchers examined the water consumption of AI systems along their entire life cycle. The study was funded by the Federal Ministry of Research, Technology and Space.
The question should be illuminated how many water resources actually need the development and use of AI. Not only was the water consumption for cooling systems examined, but also for the production of chips specializing in AI, for example.
“The water consumption of AI is often underestimated at the moment. If we do not counteract, new usage conflicts face – especially in regions that are already fighting with water shortages,” said study author Lena Hoffmann. The main problem is missing transparency, since there is no information on the actual consumption for many AI systems.
“At the moment it can hardly be quantified of the ecological footprint a single AI request-for example for ideas for a birthday present or a few lines of code-concrete,” said Teresa Zeck, a consultant at the GI and co-author of the study. “This lack of transparency makes both problem awareness and political control difficult.”
Recommendations for more sustainable AI
After their meta -analysis, the study authors are certain that AI can only contribute to coping with global challenges if it is also made sustainable and responsible. For this, an ecological transformation of the digital infrastructure is inevitable.
In order to achieve this, the direct and indirect resource consumption of AI must be considered holistically in the future. This also requires research and innovation promotion for resource and energy-efficient AI technologies that must be promoted.
For the study authors, this includes water-saving cooling technologies as well as energy and resource-saving data center infrastructure. But the location for data centers must also take place, taking into account the local water availability. In addition, a consistent circular economy must be built up in hardware production.
Also interesting:
- Also unpublished: Facebook wants to save all your photos in his cloud
- Connect Chatgpt account to WhatsApp-that’s how it works
- The best AI models in terms of data protection-and what you save
- Delete impossible: Openai has to save chatta dates permanently
The contribution study: How artificial intelligence can become more sustainable first appeared on basic thinking. Follow us too Google News and Flipboard Or subscribe to our update newsletter.
As a tech industry expert, I believe that the study on how artificial intelligence can become more sustainable is crucial and timely. AI technologies have the potential to significantly impact our environment and society, both positively and negatively. It is important for the tech industry to prioritize sustainability in the development and deployment of AI systems.
There are several ways that AI can become more sustainable. One key aspect is energy efficiency. AI systems often require substantial computing power, which can lead to high energy consumption and carbon emissions. By optimizing algorithms and hardware, and using renewable energy sources, we can reduce the environmental impact of AI.
Another important consideration is the ethical use of AI. Ensuring that AI systems are built and deployed in a way that is fair and transparent can help mitigate potential negative impacts on society and the environment. This includes addressing biases in data and algorithms, protecting privacy, and promoting accountability and oversight.
Furthermore, the tech industry can play a role in promoting sustainability by developing AI applications that address environmental challenges, such as climate change, resource management, and conservation. By leveraging AI to improve efficiency and reduce waste in various industries, we can contribute to a more sustainable future.
Overall, the study on how artificial intelligence can become more sustainable highlights the importance of considering the environmental and social implications of AI technologies. By implementing sustainable practices in the development and deployment of AI, we can harness the potential of these technologies to create a more sustainable and equitable world.
Credits