The Environmental Impact of AI, Written by and for Creative Professionals and Visual Storytellers
As an environmental studies major and filmmaker, I can't escape the environmental implications of AI. Here's run-down from what I'm seeing with AI's relationship to the environment.
If you want to support my independent reporting, subscribe to Cinema Ex Machina for free. You can also become a supporting member or you can leave a tip. 🍻 Cheers!
I plan to follow the environment closely and my first question was: for the AI that we as filmmakers use on the daily, what does that resource consumption look like?
For us, it's a bit different than the bigger AI issue and might inform the future of how resource-consumptive our tools are.
On Device Vs Cloud
From my research, Adobe, DaVinci and FCPX do most of their AI processing on-device as opposed to sending anything to the cloud. For example, if you're making an OpenAI query, that's being sent to the data centers you hear so much about. If you're masking or transcribing, that's processed on your computer in your edit bay.
Adobe
When Adobe introduced transcription to Premiere it originally sent audio to be transcribed to the cloud, now because of privacy concerns the service now runs stock on-device.
Further, from a press release in '24:
All AI features shipping with Premiere Pro now run on-device and take advantage of the CPU and GPU for hardware-accelerated performance. - Adobe
DaVinci Resolve
Using their AI Neural Engine, which runs locally and uses your system's hardware, DaVinci does all their AI work locally.
FCPX
Using Apple's Neural Engine and on-device machine learning, FCPX also does all of its AI processing locally.
No Cloud For You
As you can see, using your tools generally doesn't touch the cloud. This has other implications and the best comparison I can think of is Apple's feature: Charge iPhone with cleaner energy sources.
Apple said "we want our phones to use less energy, consume less resources because there's millions of these things connected to the grid at any one time". To solve this, they created software that charges your phone at optimal times for energy consumption or when cleaner energy sources are available.
Which brings me to my point, if all of us transcribed an interview at the same time we're still using a ton of energy collectively even if we're not working with the data centers direct.
In conclusion, energy consumption will always be an issue whether we're processing in the cloud or on-device. Let's get into it:
Energy Consumption
MIT Technology Review has a great article: "We did the math on AI’s energy footprint. Here’s the story you haven’t heard."
In 2017, AI began to change everything. Data centers started getting built with energy-intensive hardware designed for AI, which led them to double their electricity consumption by 2023. The latest reports show that 4.4% of all the energy in the US now goes toward data centers.
It’s likely that our AI footprint today is the smallest it will ever be. According to new projections published by Lawrence Berkeley National Laboratory in December, by 2028 more than half of the electricity going to data centers will be used for AI. At that point, AI alone could consume as much electricity annually as 22% of all US households.
And when talking about a new AI model that specifically generates video:
The new model uses more than 30 times more energy on each 5-second video: about 3.4 million joules, more than 700 times the energy required to generate a high-quality image. This is equivalent to riding 38 miles on an e-bike, or running a microwave for over an hour.
In short, energy use by data centers is sky-rocketing.
Climate Change
The implications for climate change are dire with this kind of resource consumption. From another MIT article:
An August 2025 analysis from Goldman Sachs Research forecasts that about 60 percent of the increasing electricity demands from data centers will be met by burning fossil fuels, increasing global carbon emissions by about 220 million tons. In comparison, driving a gas-powered car for 5,000 miles produces about 1 ton of carbon dioxide.
This is squarely where the problem lies. As consumers of energy we're going to feel this demand in our bills directly and even more directly as the climate continues to change.
A light at the end of the tunnel?
The above is pretty bad and ominous but I personally see a path which could be quite glorious. What if AI powered the green energy revolution? We're due for one with solar and wind's prices crashing, and it seems like the opportune time to re-wire the grid if we're spending the billions anyway. I don't know, just an idea from a lowly filmmaker. I guess we need to vote for the right people but that is beyond the scope of this blog.
Did you find value in this article? I worked hard to bring this issue to the fore with multiple sources and a 2000 word post. I plan to research and follow this as it develops, please consider becoming a supporter so I can continue this important work.
Water Usage
As someone who lives in the arid Western US, this is obviously a concern! Here in Bozeman, water is a legitimate concern for us:
Climate change is leading to lower yields of water in our river systems, less water and more frequent droughts, and changing the nature and timing of that water flow — early runoff, later rain in the winter months and we’re not building snowpack in the way that we’re used to.
So it's something I think and am concerned about.
In 2023, US data centers used 17 billion gallons of water for cooling. In 2028, that number is expected to reach 34-68 billion gallons/year. Creating a direct competition between AI data centers and households and agriculture.
On the other side of this, there's a pretty good case that the argument that data centers are crazy consumers of water is overblown.
From Andy Masleys The AI water issue is fake.
AI data centers use water. Like any other industry that uses water, they require careful planning. If an electric car factory opens near you, that factory may use just as much water as a data center. The factory also requires careful planning. But the idea that either the factory or AI is using an inordinate amount of water that merits any kind of boycott or national attention as a unique serious environmental issue is innumerate. Individual data centers can sometimes stress local water systems in the way other industries do, but when you use AI, you are not contributing to a significant problem for water management compared to most other things you do in your day to day life. On the national, local, and personal level, AI is barely using any water, and unless it grows 50 times faster than forecasts predict, this won’t change.
Andy talks about consumptive vs non-consumptive use, he also makes a cogent argument that a data center is no different than any other industry in water use. Last, and most importantly, he claims that when folks talk about AI water use, they're grouping a data center's water-use with the water-use by the energy source that the data center uses. Which, he's right, isn't quite the truth. To go back to the light at the end of this tunnel, renewable energy could solve this problem tomorrow and it's why I almost see the environmental catastrophe of AI as a short-term/medium-term issue.
I will be revisiting Andy's writing in a later post.
AI For Good
Artificial Intelligence (AI) can play a powerful role in supporting climate action while boosting sustainable and inclusive economic growth. However, limited research exists on the potential influence of AI on the low-carbon transition. Here we identify five areas through which AI can help build an effective response to climate threats. We estimate the potential for greenhouse gas (GHG) emissions reductions through AI applications in three key sectors—power, food, and mobility—which collectively contribute nearly half of global emissions. This is compared with the increase in data centre-related emissions generated by all AI-related activities.
They're saying in the power, food and mobility sectors which account for nearly half of all global emissions AI could reduce annual greenhouse gas emissions by 3.2 to 5.4 gigatons by 2035:
This reduction is projected to outweigh the increase in emissions from data centers due to AI activities.
Which is an insane statement! This article goes on to say:
- Optimized grid management and renewables integration in the energy sector
- Accelerating the adoptions of alternative proteins in the meat and dairy sector
- Help with public mobility efficiency
- Improving EV adoption through better battery tech and charging infrastructure
Are the potential uses that AI could help with. This to me sounds like a pitch but if AI is capable of these things, I'm on board! Let's get started!
In Conclusion, Here's What I've Seen and Recommend Thinking About
- Transparency and accountability - I'd like to see these companies go out of their way documenting and sharing their water and energy use.
- Energy efficiencies - as the nature paper talks about, use this technology to create efficiencies, fast.
- Renewable Energy - it's such an incredible opportunity for us to lead the world not only in AI but in renewable energy adoption. Vote accordingly! I think this should be the compromise for the billionaires. OK Mr. Billionaire, install your data center but you're also helping to invest in America's green, revolutionary grid.
- Regulatory oversight - we need our governments to help us regulate our environmental concerns and create an equitable playing field so we aren't footing the bill for this energy/tech revolution.
Thanks for reading, I'd love for this issue to be a continuous conversation so please email me, leave a comment, hit me up on socials with anything you've seen in this space. I'm excited to keep learning here.
