Chilean Community Runs Human‑Powered Chatbot to Expose AI’s Hidden Water Footprint

SANTIAGO, CHILE — On Saturday, about 50 residents of Quilicura, a municipality on the northern edge of Santiago, took part in an unusual technology experiment to illustrate the environmental cost of artificial intelligence: they fielded chatbot requests entirely by hand. The initiative, dubbed Quili.AI, was designed not to replicate AI automation, but to highlight the “hidden water footprint” tied to large‑scale AI data centers in the region.

For 12 hours, volunteers worked from laptops in a community center, responding manually to questions and even creating hand‑drawn illustrations on request. Despite the slower pace compared with tools like ChatGPT or Google’s Gemini, the experiment attracted more than 25,000 requests from around the world — a testament to global curiosity about how chatbots work.

One whimsical interaction involved a request for a picture of a “sloth playing in the snow.” Rather than an instantaneous AI‑generated image, a volunteer took about ten minutes to sketch a cartoon sloth gripping a snowball — a playful contrast to automated outputs.

Turning Technology Into Environmental Awareness

Organizers, led by Lorena Antiman of environmental group Corporación NGEN, said the project was never meant to reject AI outright but to spark reflection on its environmental implications. “The goal is to highlight the hidden water footprint behind AI prompting and encourage more responsible use,” Antiman said.

The initiative underscores a growing debate about the sustainability of AI infrastructure. Modern AI systems rely on massive data centers run by tech giants like Amazon, Google, and Microsoft. The powerful servers in these centers not only consume huge amounts of electricity but also use substantial quantities of water for cooling, a resource already scarce in many regions where these facilities are located.

Quilicura’s Environmental Strain

Quilicura has emerged as a hub for cloud computing infrastructure, attracting multiple data centers. But the area — like much of central Chile — has faced a prolonged decade‑long drought, exacerbating concerns about water scarcity and environmental stress. Experts say drought conditions contributed to recent wildfires and continue to affect local water supplies.

In recent years, large data center projects in the Santiago region have drawn community scrutiny and legal challenges over water use. For example, Google’s planned facilities have faced opposition and court challenges linked to water consumption, even as the company maintains some centers are highly energy‑efficient and has invested in local environmental projects.

A Different Kind of Chatbot Message

Rather than delivering instantly generated text or images, the human‑powered chatbot encouraged a slower, more mindful interaction. Volunteers answered questions drawing on local cultural knowledge — from how to make the Chilean fried pastry sopaipillas to other everyday topics — underscoring both the value and the environmental cost of AI’s speed and scale.

Antiman said the project also aimed to shift how people think about “casual prompting” — everyday use of AI tools — particularly in water‑stressed regions where the environmental toll of massive computing infrastructure is often overlooked.

Broader Environmental and Tech Debate

The Quilicura event comes as governments, researchers, and communities worldwide grapple with the environmental impact of digital technologies. Data centers are essential to modern life, powering everything from cloud storage to AI applications. Yet their energy and water demands have prompted calls for greater transparency, regulation, and investment in sustainable infrastructure.

Critics argue that relying heavily on water‑intensive cooling systems in drought‑prone regions raises ethical and ecological questions, while proponents of AI note the benefits these technologies bring to economies and societies. Initiatives like the human chatbot project seek to balance these narratives by drawing attention to often unseen costs.

Looking Ahead

As AI integration expands across industries and daily life, discussions around sustainability — including water use, carbon emissions, and local community impact — are likely to intensify. Grassroots actions like Quili.AI may help ensure these debates include voices often excluded from technology planning decisions.