Fighting Climate Change with AI
To revisit this article, select My Account, then View saved stories
Fighting Climate Change with AI
Climate change's global impact is worsening with each passing day, and society is struggling to keep pace. Is artificial intelligence the key to preparing for—and fighting against—our world's greatest threat?
Save this story for later.
Many of the core challenges—and solutions—of the climate crisis have been clear to scientists for decades. It boils down to this: The planet is warming, most likely as the direct result of increased carbon dioxide and methane emissions that began to rise sharply at the outset of the industrial revolution. The clearest solution has long been for the nations of our world to come together to develop and implement new technologies and strategies that could reduce our dependency on the fossil-fuel consumption most responsible for those emissions. And the path to get there requires that leaders around the world understand the imminent effects of climate change—so they can make informed decisions that better protect their people and places from physical and socioeconomic risks.
And yet, as a recent report on climate change argued, the world is still not doing enough to mitigate the worst effects of climate change or establish effective and equitable strategies to adapt to the changes already underway. Why is that? The data exists, but what’s lacking is a way to synthesize this vast quantity of information and expertise in a way that’s usable and actionable.
In the past few years, however, scientists working across a variety of disciplines and industries have begun to see promise in a new tool that hasn’t yet been widely applied to the climate crisis: artificial intelligence.
While the public conversation around AI has largely focused on consumer technologies like self-driving cars and smart homes, leveraging AI in the climate crisis makes sense. For decades, policymakers, scientists, and individuals have been daunted by the sheer size and complexity of the problem. Artificial intelligence, stated simply, is a tool that can turn incomprehensibly complex data sets into understandable and usable predictive models. It represents troves of potential for both climate change resilience and mitigation.
Politicians and industry leaders are beginning to understand AI’s potential in the climate fight. According to research conducted by Boston Consulting Group (BCG), 87 percent of global public- and private-sector climate and AI leaders see AI as a helpful tool in the fight against climate change. But adoption is slow. That same study found that these leaders also see clear obstacles to the use of AI for climate at their organizations, including a lack of access to AI expertise, limited availability of AI solutions, and a lack of confidence in data and analysis.
That may be set to change.
THE GREATER EFFORT TO DEPLOY AI IN THE FACE OF CLIMATE CHANGE
Over the past few years, data scientists have begun to uncover the potential of using artificial intelligence to drive new solutions to the climate crisis. In 2019, several of these scientists published a report that outlined 13 possible applications for AI . These included ideas from improving smart grids to optimizing supply chain management and developing better scheduling strategies for emergency management and response. The throughline in many of these applications was the way in which AI could be used to manage the great complexity involved in crunching carbon emissions measurement data and predicting the multifaceted impacts of mitigation solutions—both of which are not only critical to battling climate change, but also founded in more complicated data modeling than the human brain can manage.
But these applications remained theoretical. In the three years since the paper was published, several academic research projects and startup companies attempted to find practical applications for these ideas, but it has proved challenging. And what was even less reliable was finding AI applications that could play a role in averting the most devastating aspects of our future world.
BRINGING AI TO THE CURRENT CLIMATE
“You are going to be impacted by climate change. The question is, by how much?" says Hamid Maher, Managing Director and Partner in BCG’s Casablanca office.
As Maher puts it, climate analytics can play a key role in driving decision and action on adaptation by setting the true costs of inaction.
The Moroccan-born engineer found his way into climate change AI after working throughout Africa on other projects involving economic development, agriculture, and social health programs. The further he engaged in these fields, the more he discovered they all involved dealing with climate change in some capacity.
“At some point, we started to realize, project after project, that climate was a massive factor,” Maher says. “We started to leverage climate modeling progressively to inform decision-making, and we formed what has become a global team of climate scientists, data scientists, and economists working together.”
What Maher and his team recognized was that the typical decision-making frameworks around setting public policy and making investment calculations did not adequately address the ways in which the climate crisis might impact the future. Businesses could understand the cost of action—that any climate-related spending would result in smaller margins—but without a clear understanding of what the return on that investment would be. Maher and his team developed AI tools to connect the dots, creating focused models that could demonstrate how a changing climate might impact a specific city or neighborhood.
“Being able to go all the way from climate scenario,” he says, “to physical risk, direct impact, and socioeconomic impact—and derive the cost of inaction in a very robust way—is not only an analytic path that is quite big, as each location matters, but also a really complex one, as the chain effects behind the impact of a given risk in a given location are quite difficult to apprehend.”
The challenge was that the data existed, but it was fragmented. Models existed, but they were numerous and difficult to integrate. Experts in the field were out there, but they were hard to mobilize. His team focused on bringing together these various data, models, and expertise in a way that could support decision-makers in a timely manner. And as Maher and his team began to have success in creating such models, they saw the power of simply being able to quantify the future impact of climate change in a more concrete and understandable way.
Maher’s work in the field shows the wide-ranging implications for applying AI to the climate crisis. It isn’t just that AI can help us better map carbon emissions or more accurately predict the success of mitigation or adaption efforts. AI also takes the massive and nearly incomprehensible problem and boils it down into human-scaled quantitative models that can make it easier for politicians and business leaders to make informed decisions around action.
“You are a decision-maker for a country—where to start?” he asks. “You need to understand the climate change scenario. You need to understand how physical risk will evolve—and in terms of physical risks, you have so many. You have drought, extreme temperatures, rising sea levels, floods, bushfires—all those events can happen in a differentiated way from territory to territory. And that’s not it, because once you know what could happen, the second step is to understand the impact on people, on infrastructure assets, and on economic activity.”
In terms of informed decisions, Maher is acutely aware of the challenges that accompany a solution like AI. Responsible AI practices are as essential in the climate space as they are anywhere. Because AI models can be extremely resource-intensive, we must be conscientious in our usage of key resources like electricity that the AI itself requires. And AI models and datasets must be developed with climate in mind so that their outputs don’t work against themselves. But an eye for these challenges now is key to overcoming them.
SHIFTING FOCUS TO THE FUTURE
As Maher made strides in the face of climate change resilience, across the globe, his colleague Charlotte Degot was at work on using AI to power solutions for the sister piece of the climate challenge: mitigation.
Degot studied quantitative economics and finance at university, became a consultant at BCG, returned to school to study data science and artificial intelligence, and eventually joined BCG’s AI division. Like so many of us, Degot found the scope of the climate crisis scary and overwhelming. But after the onset of the Covid-19 pandemic in early 2020, Degot, like everyone else, found herself in introspective quarantine isolation—and she began to think differently about the situation. Maybe one individual couldn’t convince the world’s nations to shift their attitudes and actions around climate change, but perhaps she could apply her skills and education to the issue in her own way.
“I had a bit of a wake-up call,” Degot says. “I’m part of the most educated people in my country. I am working close to the decision-makers of the chief corporations of the world. I know a lot about tech and artificial intelligence. I’ve seen it work, and I’ve seen it deliver a massive impact on many topics. Why couldn’t it work on climate change?”
Degot established a team that has also been developing its own set of tools to help companies better model their carbon emissions so that they can make more informed decisions around mitigation strategies.
Many of the carbon abatement strategies Degot and her team identified could be implemented without an outsized impact on a company’s bottom line. And when BCG’s clients discovered this, they were eager to work on implementing the team’s mitigation roadmap. In less than a year, Degot’s team grew to more than 70 data scientists and strategists to meet growing demand for their carbon mapping and mitigation tool, CO2 AI by BCG . The toolkit allows companies to manage their end-to-end net-zero journey, identifying emissions-reducing strategies that help clients track reduction progress over time. Working across multiple industries, including oil and gas, biopharmaceuticals, retail, and consumer products, this strategy has helped clients reduce emissions by as much as 40 percent.
AI-ENHANCED HUMAN CAPACITY
Over the decades since scientists first began sounding the alarm on climate change, it has been difficult to prompt companies and individuals to take concrete action to mitigate its worst effects. Degot believes there is a simple reason for that: Human brains tend to prioritize short-term crisis mitigation over long-term considerations. It’s not that people don’t want to help solve the climate crisis—it’s that our brains don’t seem wired to do so.
“If you had the risk of losing consumers tomorrow, your brain is going to value that and prioritize that,” Degot says. “If it was easy and feasible, everyone would have a positive impact on the planet. The issue is more the complexity and the difficulty it can put on your business.”
That’s what the tools in BCG’s AI division do best: They use machine learning to rewire human brains around climate change action. They make short-term solutions out of long-term challenges. They take the uncertainty of potential economic costs of climate change and lay out a predictive framework that allows companies to make more accurate cost-benefit decisions around action. Maher sees even further potential in using AI tools to help better inform climate-minded choices on an individual basis.
“By design, our society is pushing us to consume more and more; to optimize on money and satisfy our desire to be fast, to be first, and to be quick,” Maher says. “It is a lot to ask individuals to actually behave against those incentives that are structurally anchored in our society without changing the way we push information and suggestions to them. So I believe AI has a big role to play here.”
The idea of using AI to nudge more climate-friendly behavior might seem strange at first, but Maher points to examples of where it is already being implemented ethically. Maps software, for example, could roll out tools that will allow you to plan a trip and search not only for the fastest or most efficient routes, but also the most environmentally sensitive ones. The promise of AI, Maher says, is not that it takes over our decision-making processes, but rather offers new tools for making more informed decisions.
Degot’s work with artificial intelligence and climate change has revealed another characteristic of human behavior. Just as human brains are better at solving short-term problems than confronting long-term challenges, they also tend to minimize the capacity of individuals to have an impact on local or global challenges. And yet, in less than a year, Degot and her team proved that a handful of individuals can have an outsized impact on the world’s largest and most daunting challenge. Yes, they got there by leveraging artificial intelligence—but their efforts also drew from the bottomless capacity of human hope.
This article was produced by WIRED Brand Lab for Boston Consulting Group.
More from WIRED