There is a dark side of AI. At a time when we are trying to mitigate climate crisis, a dramatic increase in AI and ChatGPT data centers are contributing heavily to climate crisis. That contribution will increase dramatically, because more and larger data centers are in construction.
AI and ChatGPT are the shiny new toys that are fascinating many people. Users say it makes their work much easier and gives them the ability to accomplish so much more in less time. But let’s look below the surface and flashy packaging.
AI has advanced faster than the morality of how it’s been implemented. It’s like the Wild Wild West out there in dataland. We need to think beyond what AI can do for us and think about what it is doing to the planet. Using AI will wipe out the benefits of any lifestyle changes we make. AI in its current form accelerates climate crisis, puts us all at risk for dwindling water supplies, contributes to social injustice, creates noise pollution that negatively impacts health, and increases our electric bills.
This summer we experienced high winds, downpours, flood warnings, tornado warnings, downed trees, extreme heat days, high humidity, and greater demands on the grid. This is our new normal brought to us by climate crisis. And AI is a contributing factor.
Climate crisis does not need any help from AI. Climate is already an accelerating disaster. Fifteen billion-dollar weather disasters occurred in the US during the first half of 2025. Devastating weather events are on the increase as climate crisis progresses. Are we now to add AI as a contributing cause?
Where are AI data centers?
There are 3,955 data centers in 51 states. Want to know where they are? Here’s the map. Ninety-three are in Pennsylvania. The majority are in Pittsburgh & Philadelphia. Amazon plans to build multiple AI‑ready data centers across Pennsylvania, starting in Salem Township in Luzerne County and Falls Township in Bucks County.
New data centers are planned for 2025 in eleven states: Arizona, California, Ohio, Mississippi, Indiana, Nebraska, South Carolia, Wyoming, Louisiana, Alabama, Minnesota, and Texas. The cost ranges from $600 million to $100 billion (not a typo). Data center facilities are often massive, spanning thousands of square feet. Where will the electricity and water they will use come from? I don’t have an answer for that. How many people will be impacted by the noise? I don’t have an answer for that either.
The Problem with AI
- Excessive electricity use
- Excessive water use
- “Natural” gas and nuclear as an energy source
- Increased carbon emissions
- Noise pollution
- Air pollution
- Increased cost of electricity for residents
- Inability of the grid to keep up with demand
- Social injustice
- AI hallucinations (presents false or misleading information as fact)
- AI lies
The immense amount of data and processing power needed to train and develop AI models and algorithms, as well as the big rise in daily requests, requires huge amounts of electricity. Research is attempting to determine how much energy AI uses, but most companies are keeping their carbon emissions a secret. But AI’s climate impact goes beyond its emissions. To understand how AI is contributing to climate change, look at the way it’s being used.
The Uneven Distribution of AI’s Environmental Impacts. The training process for a single AI model can consume thousands of megawatt hours of electricity and emit hundreds of tons of carbon. AI model training also causes excessive amounts of freshwater evaporation into the atmosphere for data center heat removal. Environmental impacts will increase, as more data centers are constructed. There is a widening difference in how regions and communities are negatively affected, leading to social justice issues.
Noise Pollution. Data centers can be a significant source of noise pollution. Residents who live near data centers complain about the 24/7 noise. Prolonged exposure to high noise levels can lead to hearing loss and other negative health issues. Noise not loud enough to cause hearing damage can still cause stress, anxiety, and even heart disease when continuously exposed to it.
Sources of power: “Natural” gas, nuclear energy, wind, solar are all used for AI. “Natural” gas is the primary power source for new data centers. There has been no consideration regarding the climate impact.
As if these problems not enough, there are other concerning issues. Read IBM’s article, 10 AI Dangers and Risks and how to manage them.
How AI is being used
AI has infiltrated our lives in countless ways. Given all the downsides of AI, would you like to use it less? Do you even know when you have used it? Despite growing AI usage, 64% of Americans are unaware they are using it. If you have used any of the following, you have used AI.
- ChatGPT
- Your car’s GPS
- Online chatboxes
- Smart phone
- Face ID on your iphone
- Email and messages
- Google search
- Google maps
- Banking
- Netflix
- Spotify
- Credit cards
- FB, Instagram, Twitter, Tik Tok
- Online gaming
- Apple watch, FitBit
- Siri, Alexa, Google Assist
- Dictation on your computer
- Coursera, Duolingo
- Language translation
- Smart home devices
- Amazon purchases
- MRI, Cat scan
Would you like to know all the stats on AI use? You can find them here.
What can you do if the use of AI concerns you? For starters, perhaps use it as little as possible?
At Transition Town Greater Media, we have created a working group to thoughtfully discuss the AI predicament. We expect to release a set of AI related guidelines and initiatives following an initial meeting of this group.
Please contact me at biodiversity@ttgmPA.org if you have any other solutions or initiatives to address AI concerns.
Author: Marion Yaglinski is a co-founder of TTGM and chairs the TTGM Biodiversity Working Group.
Leave a Reply