AI’s Growing Appetite: How Much Is Enough?

By Doug Rogers
North of Opinion Media

Artificial intelligence has quickly become one of the defining technologies of our time. Hardly a day goes by without another announcement about a breakthrough, a new investment, or another company promising that AI will change the world. I use AI regularly in my writing, research, video production and creative projects. The technology is undeniably useful. It can save time, improve productivity and open doors to ideas that might otherwise take hours to develop. Yet as I read more about the massive infrastructure being built to support AI, I find myself asking a different set of questions. Questions that seem to receive far less attention than the technology itself.

One recent story involved Kevin O’Leary and a proposal connected to what could become one of the largest AI data centre developments ever constructed. The proposed Wonder Valley project in northern Alberta has been described as a development so large that it would rival or exceed many existing facilities around the world. Reports suggest the project could eventually consume electricity on a scale measured in multiple gigawatts, a figure so large that most people struggle to comprehend it. We are no longer talking about a building filled with computers. We are talking about a vast industrial complex dedicated to processing, storing and generating data around the clock, every day of the year.

That naturally raises a practical question. Where does all that power come from?

Canada already faces discussions about electrical capacity, grid reliability and the future demand created by electric vehicles, industrial expansion and population growth. Adding a facility that could consume power equivalent to several cities creates a conversation that deserves public attention. Alberta possesses enormous natural gas resources, and much of the project’s energy requirements are expected to come from that sector. Even so, the sheer scale of demand forces us to consider what future expansion might mean for energy production, transmission infrastructure and long-term sustainability.

The discussion becomes even more interesting when water enters the equation. Computers generate heat, and advanced AI processors generate extraordinary amounts of it. Keeping thousands upon thousands of servers operating efficiently requires sophisticated cooling systems. Water remains one of the most effective cooling mediums available, which means many large facilities consume significant quantities every day. In a country where drought conditions have become increasingly common across parts of the Prairies and Western Canada, water usage is no longer a secondary consideration. Communities, farmers, industries and environmental groups are all paying closer attention to how freshwater resources are allocated. Large-scale data centres are becoming part of that conversation.

Another issue that receives surprisingly little attention involves noise. Most people imagine data centres as quiet buildings where computers sit silently behind locked doors. The reality can be very different. Large facilities often require enormous cooling equipment, industrial-scale ventilation systems, backup generators, transformers and mechanical infrastructure operating continuously. Residents living near some existing data centres around the world have reported hearing a constant low-frequency hum that never completely disappears. During the day it may blend into the background. During the evening and overnight hours it can become far more noticeable.

Noise pollution has long been recognized as more than a simple annoyance. Researchers have linked prolonged exposure to persistent industrial noise with increased stress levels, sleep disruption, reduced concentration and elevated risks associated with cardiovascular health. Human beings evolved in environments that experienced periods of quiet. A constant mechanical drone, even at relatively low levels, can affect quality of life in subtle ways that accumulate over time. Communities considering new industrial developments often focus on traffic, emissions and employment opportunities. Acoustic impacts deserve equal consideration, particularly when facilities are expected to operate twenty-four hours a day.

Wildlife faces similar challenges. Animals rely heavily on sound for communication, navigation and predator awareness. Continuous industrial noise can alter migration patterns, disrupt breeding behaviour and increase stress responses in certain species. Scientists studying resource development projects have observed behavioural changes in birds, mammals and other wildlife when exposed to persistent human-generated noise. The long-term effects vary from species to species, yet the principle remains straightforward. Large industrial operations change the surrounding environment, even when those changes are difficult to see with the naked eye.

Questions surrounding consultation have also emerged in relation to the Alberta proposal. Reports suggest that some Indigenous leaders expressed concerns about how information was shared and when discussions took place. Large projects have lasting consequences for surrounding communities. Economic opportunities, environmental stewardship, employment, infrastructure improvements and land-use planning all become part of the equation. People who live closest to these developments deserve an opportunity to understand what is being proposed and how it may affect future generations. Genuine consultation strengthens projects by incorporating local knowledge and local priorities into the planning process.

Yet the question that keeps returning to my mind is much larger than any individual project.

Why are there suddenly so many data centres being built across the world?

Every month seems to bring another announcement. Massive facilities are being planned in Canada, the United States, Europe, Asia and the Middle East. Technology companies are investing hundreds of billions of dollars. Governments are supporting AI initiatives. Investors are pouring money into anything connected to artificial intelligence. The scale of construction resembles a modern-day gold rush, except the resource being pursued is computing power.

Perhaps the answer lies in the belief that AI will become as essential to society as electricity, telephones and the internet. Business leaders clearly believe demand will continue growing for many years. Governments view AI as an economic and strategic priority. Researchers see opportunities in medicine, science and engineering that were unimaginable only a decade ago. Every major organization wants access to more computing capacity because nobody wants to find themselves trailing competitors in a rapidly evolving field.

From the perspective of an ordinary citizen, however, there is still room for healthy skepticism. Humanity has experienced periods of technological enthusiasm before. Railways transformed economies. Telecommunications reshaped society. The internet changed virtually everything. Each wave of innovation created tremendous benefits while also producing consequences that became visible only after years of expansion. AI may follow a similar path. The opportunities are remarkable. The challenges deserve equal attention.

My own use of AI is fairly straightforward. I use it to help organize ideas, improve my writing, explore topics and create content. Millions of people use AI in much the same way. Looking at those everyday uses, it can be difficult to understand why so much infrastructure is required. The answer lies in the staggering amount of computing power operating behind the scenes. Every prompt, every search, every image and every request trigger countless calculations occurring within seconds. Multiply those interactions by hundreds of millions of users and the demand becomes enormous.

Even so, society would benefit from asking a few more questions before accepting unlimited expansion as inevitable. How much power will future generations require? How much water can be allocated responsibly? What impacts will communities experience? How will wildlife adapt? What happens if demand eventually grows more slowly than investors expect? These questions are neither anti-technology nor anti-progress. They are the same questions thoughtful societies have asked whenever major infrastructure projects reshape the landscape.

Artificial intelligence may very well become one of humanity’s most transformative technologies. The benefits could be extraordinary. The infrastructure supporting it deserves the same level of public scrutiny as any major industrial development. Data centres may appear futuristic, but they remain deeply connected to the physical world. They require land, energy, water, resources and communities willing to host them.

As Canadians, we should welcome innovation while maintaining a clear understanding of its costs. Progress works best when enthusiasm and common-sense travel together. AI may represent the future, yet the decisions being made today will determine how that future ultimately unfolds.