Here’s a thought to start your day. Before you worry about the water used to train a large language model, ask yourself: how much water are you wearing?
There is a growing murmur of concern that AI is thirsty. We hear stories of data centres consuming vast quantities of water for cooling, and it is easy to picture this new technology as a villain. The complaint is simple and visceral. In a world worried about drought, we cannot afford to power this digital brain.
But this perspective looks at the cost of something without ever considering its value. It is like complaining about the price of a single seed while ignoring the entire harvest it will produce. Let’s reframe the problem.
What is the water cost of a single pair of jeans? According to the people who make them, "A single pair of jeans can use up to 3,800 liters of water in its lifetime." That is the equivalent of tens of thousands of AI conversations. Think about your morning coffee. The resources required to grow the beans, process them, and manufacture the cup add up to a staggering environmental footprint.
We accept these costs because they provide clear, immediate value. Yet, the resources they consume are a sunk cost. That coffee is gone in ten minutes. Those jeans will last a few years.
Viewing AI’s resource consumption as a simple cost is a failure of imagination. It is like saying you will save water in your company by not hiring any employees. You will cut down on coffee and bathroom breaks, but you will also have no one to do the work.
AI is a workhorse. It is an engine of productivity. The energy and water it uses are not just consumed; they are invested. As experts from the Brookings Institution note, "it is likely that AI will be a GPT [general purpose technology] and thus can be expected to generate broad productivity improvement ahead."
Consider this. If a team uses an AI assistant to solve a complex logistics problem in a day, a task that previously took a month, are they not saving a month’s worth of commuting, office energy, and human effort? That single day of AI computation just saved hundreds of times the resources.
The true measure is not the water an AI uses, but the water it saves.
It is already happening. AI-powered systems are tackling our biggest resource challenges:
The UK is at a crossroads. We risk falling behind, hampered by a reluctance to invest in the very infrastructure that will drive future growth. Blocking progress based on a narrow, decontextualized view of its costs is a recipe for becoming a technological backwater.
When we do this, we are not saving water. We are choosing stagnation. We are choosing to remain inefficient. The result will not be a greener country, but a poorer one, where our ability to solve any problem diminishes.
The solution is not to push back against AI. It is to lean in and optimise it. We can use AI to design circular economies, create hyper efficient food systems, and help us live more sustainably without sacrificing our quality of life.
Let’s stop obsessing over the cost of the engine and start focusing on where we want it to take us. The real waste is not the water used to cool a server; it is the human potential we squander by refusing to use our most powerful tools.
Links:
- Water Stewardship (https://www.levistrauss.com/sustainability-report/climate/water-stewardship/), Trust rating: High, Reason: Official sustainability report from a major global brand, providing a credible and self-reported statistic on their product's water footprint., date written: 2024-09-02
- How AI-Powered Leak Detection Saved Scottsdale 9 Million Gallons Annually (https://swan-forum.com/case-studies/scottsdale-fido-case-study/), Trust rating: Medium, Reason: A specific, data-driven case study from an industry forum, demonstrating a real-world application and quantifiable result of AI in water conservation., date written: 2024-09-02
- How will AI affect productivity? (https://www.brookings.edu/articles/how-will-ai-affect-productivity/), Trust rating: High, Reason: Analysis from a highly reputable, non-partisan think tank, providing an authoritative macro-economic perspective on AI's role as a general-purpose technology., date written: 2024-05-02
- AI-driven optimization of agricultural water management for sustainability (https://www.nature.com/articles/s41598-024-76915-8), Trust rating: High, Reason: A peer-reviewed scientific paper published in a top-tier journal, offering robust evidence for AI's positive impact in a critical sector for water consumption., date written: 2024-10-28
- Four Ways AI Can Help Reduce Waste in Your Supply Chain (https://www.supplychainbrain.com/articles/39591-four-ways-ai-can-help-reduce-waste-in-your-supply-chain), Trust rating: High, Reason: An article from a respected industry publication that directly supports the claim that AI can optimize supply chains for everyday goods, reducing their overall environmental footprint., date written: 2024-05-06
Quotes:
- "A single pair of jeans can use up to 3,800 liters of water in its lifetime." from Water Stewardship (https://www.levistrauss.com/sustainability-report/climate/water-stewardship/), Trust rating: High, Reason: This quote provides a powerful, relatable data point that anchors the central analogy of the blog post., date written: 2024-09-02
- "The combined volume of water saved by FIDO AI in Scottsdale totalled more than 9,460,800 gallons a year, which would otherwise have been lost revenue for the utility." from How AI-Powered Leak Detection Saved Scottsdale 9 Million Gallons Annually (https://swan-forum.com/case-studies/scottsdale-fido-case-study/), Trust rating: Medium, Reason: This quote offers a concrete, real-world case study with a specific, impressive number, making the argument less theoretical., date written: 2024-09-02
- "Because it is relatively easy to use and adapt to many applications, it is likely that AI will be a GPT [general purpose technology] and thus can be expected to generate broad productivity improvement ahead." from How will AI affect productivity? (https://www.brookings.edu/articles/how-will-ai-affect-productivity/), Trust rating: High, Reason: This quote from a reputable institution provides the macro-economic justification for investing in AI infrastructure., date written: 2024-05-02