Artificial intelligence is already demonstrating its climate change chops, for example by analyzing satellite images to better detect and monitor methane leaks from fossil fuel infrastructure.
In 2019, the organization Climate Change AI (CCAI) published a 100-page research paper entitled Tackling Climate Change with Machine Learning. The report detailed myriad ways in which artificial intelligence (AI) can help accelerate climate action, including detecting foliage growth across power line infrastructure; providing highly accurate and localized weather forecasts; and boosting the speed of algorithms to optimize electricity supply and demand.
German-owned company Altmetric, which logs online references to research papers, has said the Climate Change AI paper is one of the top 50 most-referenced documents among almost a million preprints listed on research paper repository arXiv. Preprints are academic papers published online before they have been peer-reviewed.
pv magazine print edition
Pick up your copy of pv magazine today for an in-depth look at batteries big and small; how they’re changing and how they are changing the world. What do batteries have in store for us? The August edition of pv magazine has answers to the big questions surrounding the battery energy storage revolution.
Since the Climate Change AI study was published – by the assembly of around 30 academic researchers and young professionals interested in how machine learning can combat climate change – the group has expanded its scope to include educational outreach, a research grant program, community building, and providing policy guidance for responsible AI. The group’s initiatives – and the potential of AI to be used to address climate change – have earned two of its co-chairs, David Rolnick and Priya Donti, membership of this year’s MIT Technology Review’s “35 innovators under 35.”
For Donti, receiving that distinction was proof of broader recognition of the urgent need for action on climate change. “We need a wider variety of tools and approaches to accelerate on-the-ground action, like existing work in renewables integration and sustainable agriculture,” she said. “At Climate Change AI, we seek to catalyze and accelerate this work by supporting members of the AI and climate action communities to deploy machine learning where it can make an impact.”
The body recently launched its ‘innovation grants’ research funding program to provide financial support for studies into the intersection of climate change and AI. The $2 million grant scheme is funded by the Quadrature Climate Foundation owned by London-based investment fund Quadrature Capital; and Schmidt Futures, a philanthropic initiative founded by Wendy Schmidt and former Google CEO Eric Schmidt. The grants will provide up to $150,000 to researchers in OECD (Organization of Economic Cooperation and Development) countries. Grant recipients must create novel, public datasets and demonstrate a pathway to real-world deployment for their research.
Climate Change AI is also working with Canada’s International Development Research Centre and the Swedish International Development Agency to evaluate proposals for an AI and climate change research hub in sub-Saharan Africa. The CA$1.2 million (US$951,000) center would disburse grants in Africa as part of the four-year Artificial Intelligence for Development in Africa program which focuses on AI for climate action.
The grants program is part of the group’s broader vision to accelerate action on climate change. Much of Climate Change AI’s work focuses on bridging the gap between energy organizations interested in AI who face challenges leveraging the technology, and members of the machine learning community interested in contributing to climate action and uncertain how to start.
“We are very eager to brainstorm with power-sector organizations on how to put our programs and activities to work for them,” said Donti, who leads the group’s power and energy sector work. “If a power-sector entity is interested in better understanding how to incorporate AI, we will be glad to chat. We are also excited to build partnerships around impactful initiatives in this sector, for instance, by creating AI researcher-in-residence programs to facilitate AI knowledge sharing within the energy sector.”
Climate Change AI sees itself as a resource for those aiming to shape programs to bridge the gap between AI and climate change action. “Our broad base of expertise and engagement has given us a good understanding of the opportunities, as well as the pitfalls to watch out for in these areas,” said Donti.
To bridge the knowledge gap between the climate action and AI communities, CCAI has developed educational and capacity-building efforts. Short tutorials on machine learning and climate change are hosted on its website, and a full summer school and course series are being created.
“We recently launched our CCAI Wiki, which is a living, breathing extension of our comprehensive research paper,” said Donti. “We also have a monthly newsletter with an extensive listing of jobs and other resources, as well as an online discussion platform for community members to ask questions and share ideas.”
The group has run events for the last two years, including workshops at leading machine learning conferences – with videos hosted on the Climate Change AI website – and a side event at the COP25 UN climate change conference in Madrid in 2019. The organization also hosts a ‘happy hour’ every fortnight to foster community building.
“The goal is to tackle climate change, not get AI implemented in every corner of climate action,” added Donti, to underscore the group’s understanding of the limitations, as well as the potential of artificial intelligence.
The volunteers who make up Climate Change AI have provided policy guidance to governments and the body has also contributed to the Global Partnership on AI – an OECD initiative aiming to bridge the gap between theory and practice on responsible, trustworthy machine learning – by helping draft recommendations to facilitate responsible AI use in the fight against climate change.
Donti said: “AI has the potential to both be helpful and harmful towards action on climate change.” For example, machine learning can help oil and gas firms extract fossil fuels, help online retailers drive consumption, and improve the performance of self-driving cars, to the potential detriment of public transport use.
As part of Climate Change AI’s work with the Global Partnership on AI, the former has issued a call for more examples of responsible artificial intelligence use to combat climate change.
CCAI co-chair Donti said: “We want to be a resource to all organizations working at the intersection of climate action and AI. We are continually looking for more ways to enhance our impact and strengthen this community.”
By Dustin Zubke
This content is protected by copyright and may not be reused. If you want to cooperate with us and would like to reuse some of our content, please contact: firstname.lastname@example.org.