The enormous electricity demand created by AI data centers may have an unexpected solution: using AI itself to make the existing power grid dramatically more efficient. The central argument from Josh Wong, founder and CEO of ThinkLabs AI, is that the world does not necessarily lack electricity; instead, much of the available generation is poorly coordinated with when and where electricity is needed. AI could help utilities manage this increasingly complex system.
A major opportunity is optimizing the grid in real time. Renewable sources such as solar and wind produce electricity intermittently, while data centers require extremely reliable power around the clock. AI can forecast demand and renewable generation, coordinate batteries and other flexible resources, and identify unused capacity across transmission networks. This could allow utilities to extract more value from infrastructure that already exists rather than waiting years to build new generation and transmission.
The article's argument is especially relevant because AI data centers are becoming a major source of new electricity demand. Building enough power plants, transmission lines and substations to serve that growth can be expensive and slow. A smarter grid could effectively increase the amount of usable electricity without requiring an equivalent increase in physical infrastructure. AI therefore becomes both part of the energy problem and potentially part of the solution: the same technology driving demand for power can help optimize the system supplying it.
The bigger takeaway is that the AI-energy relationship may be less about simply building more power and more about making existing energy systems intelligent and flexible. AI could help coordinate generation, storage, transmission and consumption at a scale that traditional grid-management systems struggle to handle. But this does not eliminate the need for new generation and infrastructure—the rapid expansion of data centers will still require substantial investment. The more realistic vision is therefore a combination of more electricity + smarter grids + AI-driven optimization, rather than AI somehow solving its own energy footprint on its own.
