5 ways AI is healing the planet

To learn more, join our free online event, AI for the Earth, on 27 March.

Could AI be a death knell for the climate? By next year, the world’s data centers could use twice as much electricity as they did last year.

Generative AI – things like chatbots and image generators – is drilling holes in our carbon budget. And yet, there are countless other AI tools helping us protect forests and wildlife, improve food security and better prepare for climate disasters to come.

If we want AI fighting on our side against the climate and biodiversity crises, it’s vital that we avoid tarring all of it with the same brush.

So, from bioacoustics to early warning systems, here are five ways we can use AI to keep our planet livable.

Protecting against climate disasters

You know that feeling when you leave your house on a sunny day, expecting the weather to hold up, only to be drenched by a rain shower appearing out of nowhere?

One of AI’s biggest assets is its ability to crunch numbers and turn them into accurate predictions. That includes improving weather forecasting, which is growing ever trickier as the climate crisis makes weather patterns more unpredictable.

In Europe, a new AI-powered forecasting system has outperformed existing forecast models by up to 20 percent and can predict hurricane systems up to 12 hours earlier.

What’s more, the developer says it uses around 1,000 times less computing energy than conventional models, while also generating forecasts much more quickly. Energy companies are already using these forecasts to decide when and where to distribute power across the grid.

That predictive power applies to extreme weather, too. AI can be used to improve early warning systems to save lives during storms and floods, as well as help farmers protect their crops from drought.

Preventing deforestation

Many of the world’s most important forests are vast, sparsely populated places – which is exactly how illegal loggers and miners often get away with their crimes.

But now, forestry officials can catch them red-handed with the help of AI.

SEPAL is an open-source platform for forest and land monitoring developed by the UN Food and Agriculture Organization (FAO), which uses AI to detect changes in land cover based on satellite imagery.

This means any suspicious activity can be identified in a matter of days rather than months, enabling authorities to respond quickly enough to intervene.

Similarly, in the Philippines, conservationists have teamed up with an AI developer to improve their monitoring of mangrove forests.

Oceanus Conservation, which operates the GLFx Mindanao chapter, is combining satellite and drone imagery with field data and feeding them into algorithms to calculate biomass, map out mangrove trees and identify tree species.

Keeping wildlife – and people – safe

Many of the AI solutions that work against loggers can also be deployed to detect poachers.

For instance, the Protection Assistant for Wildlife Security (PAWS) is an AI software developed by Harvard University that uses machine learning to analyze and predict where poachers are likely to set up snares.

This data, mainly based on satellite imagery, is then used to develop risk maps and suggest key areas for rangers to patrol. PAWS is now being rolled out across more than 800 national parks in Asia, Africa, Latin America and the Caribbean.

But poachers aren’t the only threat to wildlife, which often come into conflict with humans in far less direct ways.

In December, a passenger flight crashed in Muan, South Korea, after ingesting a flock of migratory ducks into both engines, killing 179 people.

The accident highlighted the threat that large-scale infrastructure like airports pose to birds and other wildlife – and how these conflicts can turn deadly for humans, too.

Many airports are now turning to AI to keep birds and other animals away. A Swedish startup is trialing the use of AI-powered drones at airports in the U.S. and Sweden. After detecting and identifying wildlife, the system then emits the sounds of their predators to shoo them away.

Boosting food security

Weather and climate forecasts aren’t just crucial for when you’re organizing a barbecue. They’re also a lifeline for farmers – especially as the climate crisis makes growing seasons increasingly variable.

Precision agriculture is the use of big data to optimize food production, enabling farmers to maximize yields based not only on weather but also on factors like soil conditions and crop health.

In Africa, AI-based data providers like Amini are providing farmers with the tools to determine what, where and when to plant, adapt to drought and other weather-related challenges, and boost their incomes when selling produce to consumers.

AI could also help accelerate the growth of regenerative agriculture, which can boost yields and long-term soil health at the same time. One successful project in India worked with chili farmers to improve yields by 21 percent while reducing pesticide use by 9 percent.

Preserving Indigenous knowledge

So far, none of the solutions on this list have included generative AI, and for good reason: tools like ChatGPT are responsible for the bulk of AI’s rapidly growing resource use.

But we’ll make one exception.

There’s a lot of work to be done to make AI inclusive and to bridge the ‘digital divide,’ especially in the Global South, which makes this story all the more spectacular.

As we recently reported in the ThinkLandscape digest, Indigenous developers in Brazil have built their very own AI chatbot, trained on traditional knowledge and designed to preserve their people’s cultural heritage.

Want to learn more about Mura, Sateré-Mawé and Tikuna culture? Forget ChatGPT – ask Tainá.