The use of artificial intelligence (AI) in climate action is facilitating resource management, infrastructure design and urban development, but what are the implications of employing it for our most pressing environmental challenges? The聽, 鈥淎I and Sustainability: Balancing Innovation with Planetary and Social Well-being鈥� will examine the impacts of using AI to help build a more sustainable future.
In advance of the October 10 event,聽The Reporter聽spoke with its keynote speaker, Victor Galaz. He is an associate professor in political science at the Stockholm Resilience Centre (Stockholm University), and Director of the Beijer Institute鈥檚 Governance, Technology and Complexity program at the Royal Swedish Academy of Sciences. He is also the author of the forthcoming book聽Dark Machines 鈥� How Artificial Intelligence, Digitalization and Automation is Changing our Living Planet (Routledge, 2025).
The theme of your talk is 鈥淐limate Action in an Era of Dark Machines.鈥� Can you expand on the concept of Dark Machines?
I chose to use the term 鈥渄ark machines鈥� instead of simply 鈥淎I鈥�, first, because I believe it better captures the complexities of digital technologies, with machine intelligence, hardware and digitalization, and their interactions with humans, all of which are interconnected. Second, of course, it鈥檚 because these 鈥渄ark machines鈥� can do a lot of damage to both people and the planet unless we find better ways to guide their use, and the distribution of [their] risks as well as benefits.
In your 2014 book,聽Global Environmental Governance, Technology and Politics 鈥� The Anthropocene Gap, you investigate how we fail as a society to respond to environmental crises. Things have continued to evolve rapidly since its publication, and the International Union of Geological Sciences recently voted against adding the Anthropocene to the official timeline of Earth鈥檚 history. What is your perspective on how things have changed? How do these shifts coincide with the challenges we face with AI?
I found it quite intriguing, but not surprising, that the International Union chose to vote against the proposal. But I agree with those who claim this formal step really does not matter from a practical point of view. The main insight implicit in the notion of the Anthropocene is that the human enterprise is fundamentally reshaping all life on Earth, and how Earth鈥檚 systems function. AI is a technology that is likely to change these dynamics, at times in very unexpected ways. The question now is: Will it accelerate the 鈥淕reat Acceleration鈥� or support a just transition? That鈥檚 still very much an open question, and a political choice.
There are many concerns about the growing use of AI in our everyday lives. What ethical considerations arise when deploying AI for climate-change solutions?
I鈥檓 not sure 鈥渆thical鈥� is the right word for the choices we face, I鈥檇 prefer to call them governance or regulatory challenges. There is a need, as I see it, to take a step back from the 鈥淎I for Climate鈥�-hype we have been seeing lately and ask the difficult questions. Are the AI innovations we are seeing from governments and big tech really helping 鈥渦s鈥� solve urgent climate and sustainability challenges? Who should AI climate innovation be for? And in some cases, is it even wise, considering known algorithmic biases and harms, to use AI in areas like climate decision-making, local farming advice, disaster risk responses and science communication?
In your view, can AI鈥檚 capabilities be harnessed to contribute to the development and implementation of sustainable technologies or processes? If so, in which applications or areas do you see the greatest potential?
I find myself fascinated by projects that try to combine local (and sometimes Indigenous) knowledge augmented through AI and new data, to support climate adaptation and resilient livelihoods. I hope these projects pan out well, considering the benefits they can bring to communities. Projects such as these have a strong social component and put local needs at the centre instead of hoping that a massive proprietary AI-model will be able to 鈥渄o good.鈥�
What pressing message do you want to convey with your talk?
That the development and use of AI is in our hands, and not a magical technological force outside of our control. If we are wise enough, we can redirect the forces of digital technologies to contribute to a just and safe sustainability transition.
The Trottier Symposium on Sustainable Engineering, Energy, and Design will take place at the 不良研究所 Faculty Club (3450 McTavish St.) Oct. 10 at 6 p.m. The event is free and will be hosted in English.聽