These are the ways that machine learning and AI can help us adapt
It’s reshaping industries, driving cars, beating humans at Go and even determining what you see when you log-in to Netflix, but can artificial intelligence (AI) help us meet the challenge of climate change?
C
an computer intelligence help with mitigation and adaptation? According to leading researchers, it can. In a paper first published in June, Tackling Climate Change with Machine Learning, a group of academics from around the world set out the ways that AI can be applied to the fight against climate change. “Many ML [machine learning] practitioners wish to act, but are uncertain how,” they write. “On the other side, many fields have begun actively seeking input from the ML community.”
We have trained computers to analyse vast datasets more quickly and accurately than human beings, to model countless potential scenarios in order to choose the best, and even to recognise objects. As the authors argue, there are many possible applications for these AI capabilities and others.
Using AI to reduce greenhouse gas emissions
First, it’s worth noting that artificial intelligence is the process of using computers to interpret data and follow rules to accomplish tasks that we would previously have assumed to require human intelligence. For our purposes, ML is a subset of AI, though these terms can be a little fuzzy. For this article, we will use ‘AI’ to refer to all the applications in question.
Many of the AI applications identified in Tackling Climate Change with Machine Learning are focused on reducing emissions of greenhouse gases. For example, AI can consume huge amounts of data and identify patterns, which can be used for forecasting and optimising everything from electricity supply and demand, to public transport usage.
Modeling the use of electric vehicles can determine the best time to charge them, with a view to scheduling that to happen when there is less demand on the grid, such as at night, or when more renewable energy is available.
In future, AI could get there first. Photo by Jonathan Ford /Unsplash
Protecting infrastructure from extreme events
The report’s authors also note that AI has a role to play in adaptation. For example, climate change has made extreme weather events more likely. These are difficult to predict but a December 2018 study from Harvard researchers showed that AI can predict flooding patterns at a local level using past data, which could help homebuyers, insurers and guide investment in flood defences.
Extreme events will place infrastructure under stress, which will cause failures such as electrical or water outages, or transport system breakdowns, that will make problems worse. However, AI can predict breakdowns before they occur, allowing authorities to prioritise maintenance, and can detect anomalies in how systems are functioning so that failures are discovered immediately, without having to wait for complaints to come in.
There are plenty of other areas where AI’s predictive ability could be useful, such as improving the resilience of food supply chains, optimising the performance of heating and air conditioning systems, and even modelling notoriously tricky climate change effects, such as the physics of clouds or ice sheet dynamics.
The potential here is enormous – the report runs to more than 100 pages – but as the authors note, simply pointing out the potential applications of AI in mitigating or aiding adaptation to climate change is only the beginning. The authors call for collaboration. Entrepreneurs, investors and corporate leaders must adopt these tools and apply them, while public bodies including local and national governments need to “consult ML experts while planning infrastructure” and also release data that could be relevant.
“Furthermore, technology alone is not enough – technologies that would address climate change have been available for years, but have largely not been adopted at scale by society,” the report’s authors write. They conclude: “While we hope that ML will be useful in reducing the costs associated with climate action, humanity also must decide to act.”
The ideas presented in this article aim to inspire adaptation action – they are the views of the author and do not necessarily reflect those of the Global Center on Adaptation.
