Artificial intelligence (AI) could play a role in the “great” energy transition in many ways, and as advisors guide their clients through this evolving landscape, it’s crucial to understand how AI could reshape investment opportunities, particularly in revitalizing our global electricity grid.
The least likely development is that AI conjures up some new source of energy that our human brains have yet to imagine. More promising is that AI finds inefficiencies in energy systems that engineers to date have missed. Both of these possibilities should be tempered by the potential for generative AI calculations to consume more energy than they create or save.
The greatest contribution of AI may in fact be something few have considered — that it will bring into focus the diminished state of the global electricity grid. Renewable developers, proponents of electric vehicles and heat pumps, and industrial electrification have failed to consider grid load. If AI motivates governments, utilities and investors to focus on rebuilding the global electricity grid, that will be its greatest contribution to our future.
Conduct a generative search such as “Can AI unearth the mysteries of the laws of physics?” and you will get a resounding “yes.” Change the question to “Can AI discover a new energy source?” and the output bridges optimization to exploration of existing sources. What it should return is a definitive “no.”
That said, finding efficiencies in the way we produce and consume energy is valuable. AI tools already promise better routing of transmission lines, predictive wind modelling for wind turbine performance, improved leak detection in pipelines, and more predictive matching of electricity generation with demand.
Some circumspection, however, is needed. Asset operators already employ significant data gathering and analysis capabilities. Transformative AI-driven outcomes likely depend on the discovery and installation of improved sensors, leak detectors and meters, and the data they can provide. Still, investors should take note.
This gets us to AI’s biggest contribution to the energy transition: a new-found focus on the value of a reliable and robust global electricity grid.
Over the past decade, global electricity demand has grown at just over 2% per annum. So far, generation capacity and the global grid system have mostly been able to meet this growing demand. The next decade will surely test this resiliency as demand is likely to accelerate as electric vehicle charging, heat pumps and other clean technologies come online.
Generation should be able to keep up, given that investment, particularly in renewables, has been encouraging. Transmission infrastructure investment, however, has not kept pace. Arguably, the grid has become the greatest bottleneck in the energy transition. Today, almost one terawatt — Canada’s total generating capacity times 10 — of renewables projects cannot secure grid connections.
AI’s power demand will compound this conundrum. The International Energy Agency (IEA) estimated that data centres alone consumed around 460 terawatt hours of electricity in 2022 — about 1% of global demand. Generative AI searches can be 10 to 35 times more energy intensive than traditional searches. As we integrate AI more widely, attributable power consumption could soar. Current estimates project AI could eventually account for 2.5% to 6% of total global electricity generation.
Global predictions surely mask regional implications. For example, Bloomberg New Energy Finance (BNEF) suggested the Chicago area might see a 900% jump in data centre power demand, requiring the electricity output of four new nuclear reactors. In the same study, power utility Southern Co. predicted AI would increase its electricity sales by 6% per annum, with 80% of that going to data centres.
Market research firm Dell’Oro Group estimates annual global data centre capital investments to nearly double to more than $400 billion between 2022 and 2027. Considering recent announcements from dominant tech companies, these estimates could be low. Amazon is projecting $150 billion in data centre investments over the next fifteen years, while Microsoft, in combination with OpenAI, is planning a single next-generation super data centre project with a price tag of more than $100 billion.
To ensure reliable power, some jurisdictions will build their own generating capabilities. Combined-cycle natural gas turbines are an obvious choice, and solar with battery backup will be a cost-effective alternative in the right locations. AI will accelerate sales of generating equipment, but increased generation requires investment in cables, transformers, inverters, meters, and other capital goods. As things stand, power equipment sales are soaring, with some products backordered for years.
The cumulative investment opportunity is historic. The IEA argues that to reach international climate goals, a total of 80 million kilometres of grid wires must be added and refurbished by 2040. It will require increasing current annual global investment in grids to about US$600 billion by 2030 from US$300 billion, and then to $800 billion between 2040 and 2050.
Financial advisors can take an active role in educating clients about investment in AI, as well as help ensure that the hype surrounding the technology does not distract capital away from the very infrastructure that enables it to function.
Last month, The Economist claimed that the top-100 listed AI companies cumulatively added US$8 trillion in market cap over the past year. It’s an astonishing figure — greater than the combined market cap of global airlines, global pharmaceuticals, and global automobile businesses. For a technology that is mostly misunderstood and that has an unclear business model, this development could be an historic destroyer of capital.
By providing guidance and strategies that prioritize the revitalization of our global electricity grid, advisors can help steer capital toward sustainable and resilient electricity, which will be essential as the role of AI in our economies and investment portfolios grows.
John Cook is senior vice-president, portfolio manager and co-lead of the Mackenzie Greenchip Team.