From Resource Works

Massive energy levels are required to keep up with AI innovations, and Canada risks being unable to do that

Artificial Intelligence is already one of the most important technologies of our time, and its development has been pushing innovation at a breakneck pace across huge swathes of the economy. Smart assistants now operate, albeit in a limited fashion, as secretaries for those who need help in the office, while autonomous vehicle capabilities keep improving.

It is a remarkable and world-changing time.

Just as one plays a video game, turns on a light, or starts up their car, AI requires energy. To say that AI’s appetite for energy is ravenous is an understatement, and Canadian governments must understand the challenge that comes with that.

Energy shortages are a growing threat to Canada’s economic security and, yes, our standard of living. Failure to keep up with demand means importing more energy at a cost, or facing energy blackouts, in which case Canada will fall behind in far more than just AI.

New AI models are seemingly rolling out every month, especially in machine learning and generative AI. OpenAI’s ChatGPT and Google’s Bard require huge levels of computing power to work. To train ChatGPT-4, an advanced language model, consumes thousands of megawatt hours of electricity, not incomparable to the energy usage of urban centres.

A single query made to ChatGPT requires ten times the energy of making a search on Google, revealing the massive needs of AI technology. AI is not just another internet search extension or downloadable app, it is an entirely new industry.

AI models are trained and run in data centers, which are central to this energy dilemma. The sheer power consumption in data centers is ballooning, and some estimates warn that the world’s data center energy demand will surge by 160 percent by 2030.

The International Energy Agency (IEA) has reported that AI and data centers already consume 1 to 2 percent of global electricity, a figure expected only to climb as more companies embrace AI-driven technology. As much as AI is driving digital innovation, it is also consuming electricity at a rate we will have to match.

Canada’s energy security is being seriously challenged by rising demand, with or without AI. Historically, Canadians have enjoyed the fruits of abundant, cheap energy generated by hydroelectricity in BC and Quebec, or nuclear power in Ontario. Times, and weather, have unfortunately changed.

A large and growing population, electrifying economies, and the weakening of Canada’s legacy energy sources are pushing the country to its limits regarding power supply.

The current federal government wants Canada to achieve net-zero emissions by 2050, which means that electricity is going to have to double in the next 25 years. Canada is already dealing with electricity shortages, such as in British Columbia, where demand for hydroelectricity is expected to rise 15 percent over the next six years. Manitoba is projecting a shortfall by 2029, while Ontario races to put up new nuclear power plants to avert an energy crisis by 2029 as well.

AI can help Canadians craft solutions to its incoming energy problems as a valuable research aid that can help with modeling and processing data. However, that will mean more energy consumption as part of the rogue wave of energy consumption that AI innovation has created.

As evidenced by the constant developments in AI, it is obvious that the technology is going nowhere, and neither are Canada’s energy shortfalls.

If AI is going to contribute to the surge in energy demand, then it only makes sense that it becomes a vital tool in the search for solutions, and we need those solutions now.