Your Anti-AI Post Has a Carbon Footprint Too
- jdetreglode
- Jun 3
- 5 min read
The conversation around AI and environmental impact is important. But there is a strange cultural blind spot emerging: many people criticize AI for consuming water and electricity while spending hours every day on platforms powered by the exact same infrastructure. Those infrastructures are one and the same. They are also, in fact, massive industrial data centers.
Many of the loudest critiques of AI's environmental impact are being delivered through platforms that rely on the same vast network of data centers, energy consumption, and resource extraction that make AI possible in the first place.
What I find increasingly difficult to reconcile is the disconnect I see playing out across social media every day. Perhaps "hypocrisy" is too blunt a word, but there is certainly a misalignment worth examining. Many of the loudest critiques of AI's environmental impact are being delivered through platforms that rely on the same vast network of data centers, energy consumption, and resource extraction that make AI possible in the first place. Why are we pulling from the same resources that we are claiming to want to protect?
And while I won't go too deeply into it in this post, the same question could be asked of many earth-based organizations, ecological advocates, and sustainability practitioners who I see posting daily. It is becoming imperative that we begin to examine the environmental impact of our own digital habits with the same rigor that we apply to the technologies we criticize. Accountability becomes much more powerful when it includes ourselves.
I do want to clarify that I don't think AI should escape scrutiny. But I do whole-heartedly believe that the scrutiny should be intellectually honest and applied consistently. So invite you to not ignore, but rather to explore:
Social media, streaming, cloud storage, endless scrolling, autoplay video, high-resolution photo uploads, algorithmic recommendation systems, and targeted advertising all depend on enormous data center ecosystems that have been expanding for well over a decade. The internet did not suddenly become resource-intensive because of AI. AI accelerated an already existing trajectory.
Platforms like Meta Platforms, Google, TikTok, YouTube, and Netflix all rely on hyperscale data centers that consume immense amounts of electricity and water for cooling and operation.
The uncomfortable reality is that social media itself helped normalize the very infrastructure people now condemn when discussing AI. And social media has rarely been scrutinized in the same way.
For years, society enthusiastically embraced:
infinite video streaming,
autoplay feeds,
cloud backups,
algorithmic advertising,
4K content delivery,
cryptocurrency hype cycles,
and nonstop mobile engagement
All of these require gigantic server farms drawing power and water from local communities.
Data centers consume water both directly and indirectly. Directly, they use water cooling systems to prevent servers from overheating. Indirectly, they consume even more water through electricity generation, especially when powered by fossil fuels. A 2021 paper in npj Clean Water estimated U.S. data centers consumed roughly 1.7 billion liters of water per day.
According to recent reporting, large facilities can consume up to 5 million gallons of water daily.
Meta’s own reported electricity usage reached nearly 15 million megawatt-hours in 2023 which is a 34% increase year over year. Google reported consuming billions of gallons of water annually for its data center operations, with usage increasing substantially alongside AI expansion.
But here is the key point often missing from public discourse:
Social media has already been operating at industrial computational scale for years. Every TikTok scroll, Instagram Reel, autoplaying YouTube recommendation, cloud photo sync, targeted advertisement auction, and livestream depends on the same resource-heavy physical infrastructure now associated with AI. In fact, recommendation algorithms which are the core engines of modern social media are themselves forms of machine learning and AI. The distinction people make between “social media” and “AI” is often artificial. AI has quietly powered social feeds, ad targeting, facial recognition, engagement prediction, moderation systems, and content ranking for over a decade.
This creates the misalignment/contradiction: someone posting dozens of anti-AI infographics per week on social media may themselves be contributing heavily to the exact infrastructure they criticize. The irony is that social media outrage itself increases data center demand. Viral anti-AI videos, doom-scroll discourse, reaction content, algorithmic amplification, and engagement farming all require additional computation, storage, bandwidth, and cooling.
In other words, the criticism often feeds the system it condemns.
The environmental cost of digital life did not begin with ChatGPT.
That does not invalidate environmental concerns. It simply means selective outrage becomes difficult to defend. A person livestreaming in HD for hours, uploading endless video content, or spending eight hours daily feeding engagement algorithms is also participating in the expansion of data center demand. The environmental cost of digital life did not begin with ChatGPT.
To be fair, AI does appear likely to increase demand significantly beyond previous levels. Multiple recent analyses suggest AI workloads are accelerating electricity and water consumption dramatically.
But intellectually honest environmental criticism should acknowledge the following:
the broader digital economy created this infrastructure,
social media corporations normalized hyperscale computing,
consumers willingly adopted energy-intensive digital habits,
and AI is emerging from that already-established ecosystem rather than appearing from nowhere.
The issue is whether society is willing to examine the environmental cost of the entire attention economy it has enthusiastically participated in for years.
The issue is not simply “AI versus humanity.”The issue is whether society is willing to examine the environmental cost of the entire attention economy it has enthusiastically participated in for years.
Do I have constructive feedback about how we can do better? Absolutely
! I plan to explore that more deeply in a future blog. For now, I think the most important takeaway is not to point fingers, but to cultivate awareness. Every one of us participates in the technological ecosystem, whether through AI, social media, streaming services, cloud storage, or the countless digital tools woven into modern life.
If we are genuinely concerned about environmental impact, the conversation must be larger than AI alone. It requires us to acknowledge our own habits, our own consumption, and the infrastructure that supports it all. Otherwise, we risk fighting fire with fire—using platforms powered by the same resource-intensive data centers to condemn the resource-intensive data centers powering AI.
Real change begins not with selective outrage, but with honest reflection. By becoming more conscious of how we engage with technology and the demands we place on digital infrastructure, we can move beyond blame and toward meaningful solutions that address the roots of the problem rather than its latest expression.
Note on my research: I got a lot of this information from my husband who has been a software engineer for over 20 years and for better and for worse, works extensively with AI. Other articles I have read that you might want to look into are listed below.
Comments