AI And Our Planet: It’s Complicated

Artificial intelligence may be the hottest tool shaping our future, but it’s also quietly reshaping our planet. From asking a chatbot for dinner ideas to getting real-time weather updates tailored to your exact location, all those instant answers and clever tools come with a complex set of baggage where our environment is concerned.

To make matters even more confusing, AI has incredible potential to help us adapt to a whole host of climate issues, despite the high price tag of our planet's resources.

Do all of the potential benefits of using AI outweigh the costs? That's what we set out to unpack.

Weather in your inbox

Forecast Location

Sign Up

By signing up you agree to the Terms & Privacy Policy. Unsubscribe at any time.

We spoke to experts across the fields of technology, sustainability and meteorology and came away with a different understanding of the role AI has to play.

The Promise

AI is transforming what’s possible when it comes to weather and climate prediction. “It takes hours to run [forecasts] on these massive supercomputers," explained James Belanger, Vice President of Meteorology at The Weather Company. "We can take a trained AI model, and we can leverage that to produce a weather forecast in a matter of minutes."

(MORE: How AI Has Transformed Weather Forecasting)

That kind of speed isn’t just convenient; it can literally save lives. Faster forecasts mean earlier warnings for hurricanes, floods and wildfires. AI is also being used to model long-term climate trends, helping scientists understand how global temperatures and rainfall patterns may shift in the decades ahead.

At The Weather Company, Max Jacubowsky, Principal Machine Learning Engineer, is leading the team developing AI tools that can make complex data understandable to anyone. “We have an AI that can answer weather questions like, ‘Should I go golfing today?’” he explained. “What we do is figure out what’s the right weather information to pass to the AI so that it can answer the question really, really well.”

Not only can AI take tedious, data-heavy tasks (like analyzing large forecast models) and speed them up dramatically, it can also examine and highlight important context. “Instead of just getting a human-readable output, like there's a 50% chance of rain, you could essentially ask the raw forecasts … what else could happen?” notes Jacubowsky.

(MORE: How Weather Made Boston's Most Famous Home Run Possible)

All of this means we can make smarter decisions, from choosing when to plant crops to picking the safest travel days. These are powerful examples of how artificial intelligence can support the planet and help it to be more resilient to climate change.

Which brings us to ...

The Problem

As AI grows more powerful, its environmental cost is becoming harder to ignore. The same models that help us forecast storms or chat with virtual assistants require staggering amounts of energy and water to train and operate.

"It's actually kind of in some ways contributing to the climate problem we're hoping to try to solve," Belanger points out.

(MORE: Renewables Overtake Coal As World's Greatest Electricity Source)

Imagine massive warehouses, spanning the globe, guzzling up electricity and water then spitting hazardous emissions back out. Seems like a bad deal, right? Let's break it down:

The data centers that house the technology used to run artificial intelligence on a large scale use resources in several costly ways.

• Consuming massive amounts of electricity, often including the use of fossil fuels like coal and gas
• Draining the power of the surrounding communities and in some cases hiking local energy prices
• Drawing millions of gallons of water for cooling systems needed to keep the technology from overheating
• Contributing to electronic waste, as rare earth minerals are mined to make the underlying technology and rarely recycled

Sustainable computing consultant Susannah Hill says the use of water is her biggest concern: “These new AI data centers draw a million gallons of water a day and evaporate it. They don’t return it to the sewer system, so it’s not a sustainable use of water.”

“That's the equivalent of the water that a small town would use annually,” Belanger adds.

(MORE: Scientists Worry Weather Could Impact Water Availability)

There’s also the issue of location. “When you have negative impact is when you put the data center where there isn't enough water, and then that's causing issues," Jacubowsky explained.

Beyond the numbers, there are tangible community effects. "They have 24-hour security lights that are very bright. They make a noise that is at a very low frequency and vibrates through into the neighborhood. And they also have backup generators that are almost always run on diesel, and they have to run them often," Hill described. "So living near or downwind from a cluster of data centers makes your air quality worse."

The Path Forward

In the end, AI’s climate story isn’t purely good or bad. “You want to make decisions that are good for the environment," as Jacubowsky puts it, "You also can't just be naive, because in reality, people are going to cut corners and try and prioritize profit."

Still, there’s hope. Belanger says some major tech firms are already working to prioritize sustainability. "These organizations are signing power purchase agreements specifically to target how the electricity that they're using in their data centers is being sourced.” And Hill emphasizes that “the very large tech companies have every incentive to be as efficient as possible" and reduce their power use.

(MORE: The Ozone Layer Is On Track For Full Recovery)

Smaller, more specialized models, rather than massive, general-purpose ones, could also dramatically cut resource use.

Ultimately, all three experts agree: awareness and transparency are key. “People should care about the energy AI is using because it's something that we all participate in," Hill stressed. "It’s going to be underpinning everything we do online."

"If we can prioritize and work together to understand what are those sustainable practices, it's really gonna set us up well in the long hand to make that question on whether or not it's a net positive or a net negative for society, very clearly a net positive,” Belanger predicted.

For individuals, Hill says, knowledge is power. “The good news is that it’s a problem each of us has a role to play in,” she explained

AI users can dramatically limit their own carbon footprint by:

1. Using chatbots discerningly, saving your answers and avoiding repeated questions
2. Taking advantage of offline features
3. Steering clear of AI generated images or videos
4. Advocating for sustainable computing and AI practices at work

Don’t just take our word for it. We typed this story into an AI chatbot to see what it had to say. Perhaps unsurprisingly, it summed the debate up nicely:

“Whether AI is ‘worth it’ depends on how we use it and how fast we make it cleaner. Right now, the balance leans toward potentially worth it, but only if the industry transitions to renewable energy, designs more efficient models and is transparent about its footprint.

In other words: The benefits could justify the cost if we treat sustainability as a core part of AI’s evolution, not an afterthought.”

Weather.com lead editor Jenn Jordan explores how weather and climate weave through our daily lives, shape our routines and leave lasting impacts on our communities.