AI boom: How much electricity and water is in an inquiry?

AI boom: How much electricity and water is in an inquiry?

Europa, Europa - The energy consumption of artificial intelligence (AI) increasingly ensures concern among environmentalists. According to Openaai, an inquiry to the CHATGPT AI software uses as much electricity as a second operation of an oven. Openai boss Sam Altman published this information on a blog. In addition, the water consumption per request is at a fifth of a teaspoon. Despite increasing efficiency in chip and server techniques, the total current requirement will increase due to the increasing use of the technology. This leads to a huge increase in energy requirements in the coming years.

like the Tagesspiegel reports the energy consumption of data centers for AI applications in Europe 2030 expected 150 terawatt hours. This almost corresponds to a doubling of current consumption and will make a contribution of around five percent to the total European electricity consumption. It is alarming that this increase could accelerate climate change if the additional requirements should not be covered by renewable energies.

fossil fuels and nuclear energy in focus

The majority of the electricity for data centers are currently still from fossil fuels. Large tech companies such as Microsoft, Google and Amazon therefore rely on nuclear energy to cover the energy requirement without increasing carbon dioxide emissions. Google has already signed a contract for the purchase of small modular nuclear reactors to cover its energy requirements, while the first power plant is to go into operation by 2030

An analysis of the eco-institute on behalf of Greenpeace Germany shows that the electricity consumption of AI calculating centers from 50 billion kWh in 2023 could increase to over 550 billion kWh in 2030. This forecast suggests that greenhouse gas emissions from 212 million tons in 2023 will also increase to 355 million tons in 2030. The increasing water consumption to cool this server will also be remarkable and is expected to increase to 664 billion liters.

Need to effectiveness and efficiency

The increased need for energy and raw materials entails considerable ecological challenges. Up to five million tons of electronic scrap could result from the construction of new data centers. Despite the undeniable opportunities that AI offers to improve energy efficiency in sectors such as agriculture and energy supply, it should be noted that their use can also have negative effects. This was documented by numerous studies and reports, such as "https://www.ingenieur.de/technik/kuenstliche-intelligenz/ki-boom-gefaehrdet-klimaschtz-dieben-fähren-ähren-febedarf-in-hohe/" class = "source_3" , illustrates.

A comprehensive assessment of the environmental impact of AI is essential. According to the report, binding transparency requirements and accounting obligations for data centers as well as the development of an efficiency label are required for their ecological footprint. The integration of data centers into renewable energy and heating networks could also be a step in the right direction. The goal must be to close the gap between technological progress and environmental protection.

The debate about the balance between AI and sustainable development is at the beginning. Technological advances do not automatically guarantee a better environmental balance. Therefore, a conscious political and social framework is necessary to use the potential of the AI ​​for climate protection and at the same time control the associated risks.

Overall, these developments show that the need for action is enormous to master the challenges that the AI ​​boom entails and to comply with the climate goals.