Recently, we have seen an interesting post go viral on social media. It’s effectively a claim that says golf courses in the United States waste around 30X more water per annum than AI data centers. The reason for the post’s virality is arguably down to the fact that so much effort has gone into highlighting the water wasted by AI data centers over the last couple of years, particularly after the launch of ChatGPT and other consumer-facing chatbots.
Now, the claim is a bold one, and we can’t fully verify it, but it is enough to say that it does hold water. Some estimates suggest that a standard golf course may require 1-2 million liters of water per week, whereas a large – and some are very large – data centers will use a fraction of that – perhaps around 4-5% – per week. Again, though it depends on a variety of factors, ranging from the size of the data center to the location of the golf course.
Contents
Perspective Is Needed for AI and Digital Energy Usage
The above certainly does not mean data centers are good for the environment, nor does it mean that golf courses are bad. It is, rather, a matter of perspective. Many people are angry about AI’s impact, and its future impact, on the environment, yet you probably haven’t seen too many protests about water usage at your local golf course recently.
As we enter a future where technology continues to play a more influential role, it is important to keep that perspective in mind. For example, we do know it requires a lot of energy to power a computer, but let’s say we consider the energy usage to send an email. A simple email generates approximately 0.3g of carbon dioxide. But then compare that to the energy and carbon emissions from sending a letter by post. You have to factor in several truck journeys, the power required to operate a mail sorting center, and so on. The same goes for a variety of online activities. Think about playing slot games online and then compare the massive amounts of energy to power a casino.
Humans weren’t made to be digital recluses
There are so many examples where digital is better than the in-person activity. However, that doesn’t mean that we should all be meeting on Zoom calls rather than driving to meet our friends in a restaurant or café. Humans were meant for physical connections, and the recent pandemic and its aftermath reminded everyone of that fact.
If you recall the early 2020s, there was an almost panic-driven rush to announce that WFH (work from home) was the future. Many of the Big Tech companies made the call, claiming that the option would be there for employees in perpetuity. We were told that the pandemic was the catalyst for the move, showing that it could be done rather than the root cause. It would save money and, in all likelihood, be better for the environment.
Yet, the tune quickly changed, both from employees and employers. Many of the Big Tech companies reversed those calls to offer WFH to their staff. Yes, it still exists and hybrid models of working do too, but it is also clear that there was a rejection of remoteness. Indeed, there was evidence that those employees who did go to the office were more likely to be promoted, as they were the ones able to create interpersonal connections with their management.
Questions about “pointless AI”
As we said, it is all about perspective. But if we get back to AI, we can also say there is an argument that there must be a point to it. For instance, if people are using AI – burning through energy, wasting water, etc., – to create pointless images and videos, which is now known under the tagline of “AI slop,” then you can argue that it doesn’t matter how many times “worse” a golf club is; you’re hurting the environment for no good reason. Yet, if that AI tool was being utilized towards a goal like curing cancer, most of us would agree that the energy usage was worth it.
<iframe width=”560″ height=”315″ src=”https://www.youtube.com/embed/b0C56yqIkbk?si=wAjd4imxyu3iAi_p” title=”YouTube video player” frameborder=”0″ allow=”accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share” referrerpolicy=”strict-origin-when-cross-origin” allowfullscreen></iframe>
And that essentially is the rub: We started this piece talking about how AI has received a lot of criticism for its energy and water usage – some of which is wholly merited. Yet, the industry needs to give a more compelling reason to make us support the continued building of those power-hungry data centers. At the moment, all the general public sees is AI being used to take (human) jobs in the name of productivity, while being used to create slop on the internet. There can be give and take when it comes to preventing climate catastrophe, but right now, it seems like there is more take.
It feels like there will be a reckoning at some point. At the time of writing, Meta, which made its own missteps with misjudging the desire for physical human connections with its disastrous metaverse project, has just announced the layoffs of 8,000 employees, citing cost-cutting measures as it pursues AI investment. Yet, aside from the human cost of those layoffs, people are asking, to what end is this AI investment? What’s the point of it? Meta is hardly going to pivot into medicine, so we can take curing cancer off the books, so it largely lies with ‘improving’ Facebook, Instagram, and the rest of its family of social media apps. Those improvements would likely be in the areas of more efficient content moderation, better algorithms (so you can be targeted with more personalized content), and better targeting of adverts (so Meta can make more money).
But is that in itself a good reason to burn billions in the pursuit of AI? Is it a justification for the destruction of clean water supplies to fuel data centers? AI is deeply unpopular with the general public. A study published in CNET in March of this year claimed only 26% of Americans of voting age viewed the technology positively. Not all of that negativity is due to its impact on the environment, and some of it may come from the perceived threat to jobs, but we find it difficult to believe that our perspectives will change on the matter until someone demonstrates that its impact – the financial and environmental costs – is a case of the ends justifying the means. We aren’t there yet.