Are Everyday People Generating Images to Blame for the Uptick in Data Centers?
There are really smart people who believe that data centers are being built so that “people can generate images.” That is not true.
Telling everyday people not to use AI is not likely to make a major dent in AI-related electricity use by itself.
The big electricity demand is coming from:
- massive corporate data centers
- model training and large-scale deployment
- AI has long been built into search, ads, cloud software, phones, workplace tools, surveillance, finance, logistics, weapons systems, health systems, and government systems. It “seems” like it all ‘happened so fast,” but the truth is this has been coming for decades. Everyday, people have just been shown more now.
- companies running AI whether ordinary people “opt out” or not
The International Energy Agency estimates that global data center electricity consumption was about 415 TWh in 2024, around 1.5% of global electricity use, and projects data center demand to roughly double by 2030. AI-focused data centers are expected to grow even faster.
So yes, individual consumer choices and behaviors matter, but they are not the largest lever. Much the same way that consumers have melting straws in their drinks, as if corporations, private jets, and yachts aren’t the problem.
The biggest levers are:
Utility regulation.
Who approves these facilities? Who pays for new grid upgrades? Are local households subsidizing corporate expansion?
Energy sourcing.
Are data centers powered by renewables, nuclear, gas, coal, or behind-the-meter fossil fuel plants?
Water use and local impact.
Are communities being drained, heated, priced out, or politically bypassed?
Corporate accountability.
Are companies required to report energy use, water use, emissions, and community impact honestly?
Efficiency standards.
Are companies being pushed to reduce wasteful compute, not just expand endlessly?
That is why I find the “ordinary people should stop using AI” argument incomplete.
It can easily become another version of telling working people to carry the moral burden while corporations keep building the machine. Even further, schools have AI as a part of their curriculum and degrees. Students will have to complete homework and learn new skills.
Same pattern as:
“You recycle more,” while industry pollutes.
“You turn off your lights,” while corporate buildings blaze all night.
“You stop using plastic straws,” while massive supply chains flood the world with plastic.
“You don’t use AI,” while billion-dollar firms automate entire departments.
The better argument is:
Use AI in ways that add and enhance. Learn AI skills because the change is here and the skills are beginning to be a requirement in some jobs and fields.
But regulate AI infrastructure aggressively.
Everyday people learning AI are not the main driver of the energy crisis. They are often the people most at risk of being left behind if they do not learn it.
A grounded position would be:
“Do not shame ordinary people for learning tools that institutions are already using. Put the pressure where the power is: data centers, utilities, tech companies, regulators, and public officials.”
That is the cleaner target.
And it lets people hold two truths at once:
AI can help everyday people gain skill, access, income, language support, disability support, business tools, and creative power.
AI infrastructure can also harm communities if it expands without limits, transparency, or public accountability.
Both are true. This is the same issue around energy consumption that has been debated for decades or even longer.
The answer is not to keep everyday people away from AI. The answer is to keep powerful institutions from making everyday people pay the hidden bill.
