May 22, 2026 - 2 min

AI needs building blocks: the investment boom happening behind the scenes

The real driving force behind the artificial intelligence cycle isn't the models themselves, but the infrastructure that supports them. Data centers, power grids, and cooling systems have become the new battleground for big tech companies—and an opportunity that few investors are fully recognizing.

Share

When people talk about artificial intelligence, the conversation usually revolves around models, algorithms, and competition among tech giants. But behind every query to a language model, every generated image, or every automated prediction, there is something much more tangible: servers, cables, electricity, and physical space. The infrastructure that makes AI possible has become a strategic resource, and demand is growing faster than the capacity to build it. 

The scale of the upcoming investment cycle is difficult to gauge. According to McKinsey, we are facing one of the largest industrial rollouts in decades. The most recent estimates, compiled by the World Economic Forum, project that global investment in data centers could exceed $7 trillion by 2030, driven largely by the widespread adoption of AI and the growth of cloud computing. It’s not just about building more facilities: the challenge includes the electrical equipment, cooling systems, and high-speed connectivity that each center requires. 

One of the most visible—and least anticipated—effects is the impact on energy systems. Goldman Sachs projects that electricity demand associated with data centers could increase by nearly 165% by 2030. To put this in perspective: data centers are already the primary driver of growth in global electricity consumption during this era of digital expansion, surpassing traditional manufacturing. This is driving urgent investments in energy generation, transmission, and storage, including renewed interest in sources such as small-scale nuclear power. 

This phenomenon is reshaping the AI value chain. The benefits of this cycle are not limited to big tech companies: electric infrastructure firms, semiconductor manufacturers, fiber-optic providers, and data center operators are capturing an increasing share of spending. In practice, investing in AI today can mean betting on either a language model or the company that builds the facilities that power it. 

The challenge ahead is not one of vision, but of execution. The speed of construction, the availability of energy, and the ability to coordinate very different sectors have become critical factors. The fact that Google and Blackstone recently announced a partnership to develop joint data center capacity—as reported by Reuters—is a clear sign: competition for infrastructure is no longer a secondary concern. It is the linchpin of the cycle. 

 

Fynsa

 

Sources: McKinsey & Company · World Economic Forum · Goldman Sachs · Reuters