ai uses energy.
so do refrigerators, hospitals, search engines, streaming video, christmas lights, crypto mining, air conditioning, and the entire human habit of asking a server somewhere to show us a dancing raccoon in 4k.
that does not mean ai is killing the planet.
it means ai is part of the same larger question every technology eventually has to answer:
what does this cost, what does it replace, and who benefits?
the panic shape
a lot of arguments about ai and the environment skip directly from:
- ai uses electricity
- data centers use water
- tech companies are building more infrastructure
to:
- therefore ai is uniquely evil
- therefore using ai is morally equivalent to setting a wetland on fire
- therefore everyone who asks a chatbot for help writing an email is personally feeding the doom machine
that leap is doing a lot of sweaty little backflips.
the missing context
ai has environmental costs, but those costs need context.
questions that matter:
- what energy source powers the data center?
- what workload is ai replacing?
- how many people are using the same infrastructure?
- is ai reducing other waste somewhere else?
- is the output useful, frivolous, exploitative, or genuinely helpful?
- are companies being transparent about water, power, and emissions?
without those details, “ai is killing the planet” becomes less of an analysis and more of a panic sticker slapped onto a complicated machine.
not all ai use is the same
there is a difference between:
- using ai to summarize medical paperwork
- using ai to generate 900 disposable spam articles
- using ai to help disabled people navigate inaccessible systems
- using ai to make a fake influencer sell protein sludge
- using ai to troubleshoot code instead of spending three hours rage-googling
- using ai to mass-produce garbage content because the internet apparently needed more beige oatmeal
the tool matters.
the use matters.
the scale matters.
the incentives matter.
the real problem is not “ai exists”
the real problems are things like:
- unchecked corporate growth
- fossil-fuel-powered infrastructure
- wasteful content farms
- planned obsolescence
- opaque data-center reporting
- companies externalizing costs
- pretending “innovation” automatically means “good”
ai is not exempt from criticism.
but blaming ai as if it is the singular villain is too neat. too convenient. too emotionally satisfying.
the planet is not being murdered by one spooky autocomplete box.
it is being strained by systems that reward endless extraction, endless consumption, and endless growth without accountability.
ai can become part of that.
ai can also be used against that.
the better argument
instead of saying:
ai is killing the planet
say:
ai has real environmental costs, and companies should be transparent about energy use, water use, emissions, and whether the technology is being used for meaningful work or disposable garbage.
that is stronger.
that is harder to dismiss.
that does not require pretending every ai user is personally clubbing a polar bear with a server rack.
where i land
i do not think ai is innocent.
i also do not think ai is uniquely monstrous.
it is a powerful technology being plugged into an already wasteful world.
the question is not whether ai should exist.
the question is whether it will be used thoughtfully, regulated honestly, powered responsibly, and kept from becoming another machine that turns human attention, labor, water, and electricity into corporate confetti.
tiny thesis
ai is not killing the planet.
bad incentives are.
ai is just one more place where those incentives show their teeth.