By now you’ve probably heard about Chat GPT writing students’ essays, or the Bing chatbot declaring love to journalists, and a host of other AI tools and services that are pushing boundaries and going far beyond what was previously thought possible of computers.
Yep, we’ve officially entered the age of artificial intelligence. In fact many believe that this technology is as game changing for humanity as the internet, leading some AI experts to call for an Intergovernmental Panel for Artificial Intelligence to ensure that governments work together to regulate and manage the risks.
At the core of recent artificial intelligence innovations is not a beating heart or organic neural pathways, but huge datasets. These systems can analyze vast amounts of data and make predictions without ever getting fatigued, making them useful for scientists and researchers.
In the last decades much research has been done on the potential applications of AI to climate solutions, such as mapping and monitoring forests to better protect them from deforestation and wildfires.
It’s not without a catch, though. Training an AI on such large amounts of data, even to help us to better understand complex ecosystems or to apply to climate solutions like renewable energy, uses a huge amount of computational power which comes with its own environmental costs.
In fact, one study found that training a single AI releases five times the emissions that an average car would emit during its lifetime, thereby adding to the already substantial carbon footprint created by internet technologies. According to computer scientist Benjamin Kanding, this could be reduced by training AI models in countries like Estonia and Sweden where there is a greater abundance of green energy.
At the end of the day technology alone is not enough. AI is no magic bullet for solving climate change, and a global movement towards climate action remains urgently needed. As the authors of one study on AI and climate change note, many technological tools useful in addressing climate change have been available for years already, but have yet to be adopted at scale.
In this article, we're exploring some ways in which AI could be applied to tackle climate change.
Making renewable energy grids smarter
According to a report by AI researchers called Tackling Climate Change with Machine Learning, AI could help speed up the transition to renewable energy in many ways, for example by helping to assess supply and demand.
Since the wind doesn’t always blow and the sun doesn’t always shine, energy grids employing the help of AI can make more accurate predictions by using historical weather data and adapt operations accordingly.
It can also help on the demand side. For example if it’s a rainy day and solar cells are unable to harvest sunlight, AI could shift to look for other energy sources that will guarantee an unbroken flow of energy. This will ensure that there is always enough on-hand energy to power the homes and businesses that are relying on it.
Reducing food waste
Reducing food waste by 50% by 2030 is one of the UN Sustainable Development Goals, and there are already a few examples of AI being applied in the food industry to minimize overproduction and food waste.
Some goods can be more efficiently produced using data-based algorithms that give intelligent information on supply and demand. This means that producers will only produce as much food as is likely to be needed, without as much surplus as is currently the norm.
You know how products in the supermarket can often be slashed in price right before they are about to pass their sell-by date? Using AI-powered tools, supermarkets could make “dynamic price adjustments” based on data like customer behavior, inventory levels and expiry dates. This means that food is more likely to be bought before its sell-by date, and less likely to go to waste.
Managing forests and reducing deforestation
Mapping and monitoring forests using AI can help us understand how to properly manage and protect them, especially in regards to wildfires and deforestation.
Instead of manually counting trees or measuring forest density, more companies and research labs have started to use satellite imagery combined with machine learning to better understand the layout of forests.
When it comes to out-of-control wildfires, using data on how fires spread and tools to evaluate regions that are more at risk, firefighters can perform controlled burns and thin select areas to prevent progression of fires.
Scaling sustainable agriculture
Farmers can use AI for methods such as precision agriculture; they can monitor crop moisture, soil composition, and temperature in growing areas, enabling farmers to increase their yields by learning how to take care of their crops and determine the ideal amount of water or fertilizer to use.
One recent study used AI to track bees and pollinating insects to build a picture of the behavior of pollinators and increase food production.
A 2019 report from Wildlabs.net identified AI as one of the top three emerging technologies in conservation, helping protect species around the world, supporting the work of scientists, researchers and rangers; from anti-poaching patrols to monitoring species.
The usefulness to scientists comes from the fact that AI can process huge amounts of data that would otherwise take an academic or researcher months to go through. For example, according to an article in Nature journal, Ruth Oliver, an ecologist at Yale University, used an AI tool to analyze 1,200 hours of audio from the Arctic and used that data to estimate when songbirds arrived in the Arctic, and how the environmental conditions in the Arctic affect breeding patterns. Data like this can be used to forecast changes in bird migration in the face of climate change.