Artificial intelligence is indisputably going to become a part of our lives. Tech giants like Microsoft, Google and Amazon are betting big on it, integrating it into everything from search engines to digital tools. How to understand the impact on society of this new reality? A recent study has drawn attention to the environmental cost of these models, and particularly to their water footprint. Although it might seem like a side issue, it is not. It is known that water will be the cause of the next global conflicts and, in fact, there are already tensions around water resources within the countries themselves. As the drought spreads, there will be a geopolitical escalation for a fundamental resource for life, but also for technological competition. It must enter the equation.
The sustainable perspective is essential when developing technology. The neural models like GPT-3 and GPT-4 that are behind the well-known chat rooms, along with other technologies like cryptocurrencies, have been criticized for their carbon emissions. But what about the water? ChatGPT needs to drink half a liter of water to have a conversation of 50 questions. Let’s do the numbers with the number of users you have.
Comparisons with other industries highlight the magnitude of what we are talking about. State-of-the-art data centers in the US have consumed 700,000 liters of clean fresh water for training. Enough to produce 320 Tesla cars. And it could be worse if it were in the centers of Asia, because there consumption would directly triple. Beef production and jean manufacturing also have high consumption, but their footprints take into account the entire life cycle and include a large portion of non-potable water. Adding in the water consumption associated with manufacturing and transporting the AI servers, their global water footprint could increase 10 times. But not all is lost. Once you are aware of this element, you can optimize consumption taking into account system load and geography.
It would be something very similar to how we reduce our electricity bills, playing with the prices during the night or off-peak hours. Interestingly though, carbon reduction and water conservation can sometimes be in tension. For example, in California there is a high production of solar energy towards noon, which gives rise to the hours of greatest carbon savings. However, the outside temperature is also high around noon, resulting in the worst water efficiency. Thus, if we only take into account the reduction of the carbon footprint, we can end up with a greater consumption of water. On the contrary, if we intend to reduce the water footprint, we could increase the carbon footprint due to less available solar energy.
Legislation and large producers of artificial intelligence should urgently incorporate sustainability into the catalog of necessary conditions for technological development. And especially with a resource as scarce as water. If technological warfare (the race between nations and companies to achieve mastery of technology) intersects with water wars (conflicts over scarce water resources) a new battlefield will emerge, in a world that is already sufficiently in crisis to add new elements of conflict between different actors.
Water is becoming a scarce commodity and competition for access to this vital resource is likely to intensify as the effects of climate change worsen. If technology companies continue to expand their data centers across the globe, it is essential that water use considerations be integrated into the site selection process.
Locating data centers in regions where water is scarce can contribute to local water stress. Even if these centers are built with state-of-the-art water recycling systems, the large volume of water required to keep them operational could be detrimental in drought-stricken areas. Also, the water used in these data centers has to be clean fresh water. In a country like Spain, which already suffers from water scarcity, diverting this valuable resource could cause unwanted effects between the different uses of water, such as agriculture or consumption.
This only points out to us the interconnected nature of our global challenges. It forces us to think more holistically, recognizing that our pursuit of technological advances must not come at the expense of our planet’s resources. As we navigate this new battlefield, success will be measured not only by technological capabilities, but also by a commitment to sustainability and the equitable distribution of resources. Why not start a new race to reduce consumption and environmental impact? Almost 3 billion people on the planet they suffer water stress; and that, before the revolution of the new generative artificial intelligence. The future of technology must not only be more innovative, but also sustainable. After all, what good is artificial intelligence if it endangers the very resources we need to survive?