Stop drinking water! 

Sustainability lab launches provocative campaign highlighting AI’s hidden environmental costs 

In the halls of the Faculty of Information’s Bissell Building, and as far away as Silicon Valley and Melbourne, a series of unusual posters are sparking conversation and turning heads. The “Save the AI” project, created by Professor Christoph Becker’s Just Sustainability Design (JSD) Lab, uses irony, humour and data to highlight the significant, and often overlooked, environmental and social impact of generative AI. 

With tongue-in-cheek messages that ask people to conserve water and electricity “for the AI,” the campaign draws attention to the vast resources required to power large-scale data centres and machine learning systems.  

The campaign started as a joke, asking people to stop drinking water because AI needed it. Then the team at JSD Lab expanded the idea to electricity, air, coal, and other resources consumed by the use of AI. 

The campaign is not about making anyone feel guilty, say the students behind it. Its humour is designed to bridge what researchers call psychological distance, the gap between our everyday lives and the seemingly faraway effects of systems like data centres, used to power AI. “Humour can open space for critical reflection,” says Han Qiao, a PhD student at the Faculty of Information. “People laugh first, but then they start to think, and that’s when the real engagement begins.” 

Since the campaign began back in February, it has been translated into nine languages, as people around the world join up. New research and data is continually added to the project’s website.  

“It’s not only about data visualization, but visceralization,” says Qiao, explaining that visceralization is an approach that helps people feel data rather than just see it. Her research explores how our interactions with data shape our relationships to physical spaces and our collective imaginations of urban futures. “We want people to experience the weight of these numbers and to feel what they mean,” she says. 

The researchers emphasize that the campaign aims to target the narratives devised by big tech companies rather than shifting the blame to individual users. “The goal is definitely not to shame people,” Qiao says. “It’s about building awareness.” 

“While human machine interaction, machine learning, and automation are promising fields of research and creation, we need to think about who these technologies are ultimately benefiting,” says Cristian Velasquez, Communications and Engagement Lead for the JSD lab and a Master of Information student. He cautions against the current state of AI development and implementation, pointing out that it uses massive amounts of resources and depletes already fragile ecosystems and communities.  
 
The research team hopes the campaign will encourage people to recognize that using technology should be a choice, not a default. Along with spreading the words by translating the posters into multiple languages and sharing them globally, the JSD Lab is now focusing on what comes after awareness, supporting initiatives such as Resisting the AI and Big Tech in Higher Education, a collaborative talk organized with student groups across the world, which will take place next week. 

“It’s inspiring to see students and researchers taking concrete action together,” says Velasquez.  “We hope more people can join the conversation.” With Generative AI disrupting learning and undermining trust in the integrity of academic work, as well as exacerbating ecological crises, Velazquez says we need to ask, “What can we do? How do we resist?”