Crusoe Energy's CEO explained how AI is changing the economics of data centers and the computing market

Why the future of AI depends on energy

The development of artificial intelligence requires not only more powerful models but also large-scale physical infrastructure. Modern AI systems require data centers, computing equipment, and stable power supply. As demand for AI grows, access to energy is becoming one of the main constraints on industry expansion.

In the future, AI infrastructure will develop not only around major technology hubs. Regions with significant energy resources and opportunities for deploying computing capacity will gain increasing importance.

This changes the approach to building data centers. Now, it is important for companies to consider not only the cost of equipment and the location of facilities, but also how quickly they can obtain the necessary energy and launch infrastructure.

Why companies are moving toward vertical integration

Another industry problem is supply chain constraints. Even with funding, companies may face shortages of electricity, equipment, and specialists needed to build and operate data centers.

One solution is vertical integration—an approach where a company controls several stages of product or service creation. In the case of AI infrastructure, this can include power supply, data centers, and computing capacity.

This model allows for better coordination of infrastructure development, reduces dependence on external suppliers, and enables faster responses to changing demand.

AI infrastructure should also be viewed through the lens of raw material markets, particularly the oil and gas industry. In this model, efficiency depends on managing the entire value chain, not just individual technological solutions.

How data centers affect regional economies

The construction of large data centers often raises concerns among local residents. Key concerns include high electricity consumption, strain on infrastructure, and environmental impacts.

However, such facilities can bring economic benefits to regions. They create demand for construction, engineering services, and technical specialists, and contribute to increased tax revenues.

Additional energy capacity created to serve data centers can, under certain conditions, positively affect electricity availability for other consumers as well.

At the same time, industry development requires not only engineers and developers. Building and maintaining infrastructure requires electricians, fiber optic network specialists, construction workers, and other professionals. Thus, the expansion of AI infrastructure creates demand for a wide range of technical professions.

How the GPU economy is changing

A separate topic of discussion is the cost of graphics processing units (GPUs) and the economics of computing. As AI develops, companies must consider not only the initial price of equipment but also how effectively it is used throughout its lifecycle.

Depreciation of GPUs plays an important role—the gradual allocation of equipment cost based on its usage and lifespan. The economic value of computing systems depends on performance, utilization, and how quickly equipment requirements change.

For companies investing in large computing capacity, it becomes crucial to understand how long equipment will remain in demand and generate revenue.

This is especially relevant against the backdrop of rapid AI model development and changing ways of using them. Solutions that provide a competitive advantage today may eventually give way to more efficient technologies.

Why inference is becoming increasingly important

As the AI market develops, the nature of computing load is also changing. In the early stages, much attention was paid to training models, but as they become more widespread, the importance of inference—the process by which a trained model processes requests and generates responses—grows.

For such systems to work, it is important not only to have computing power but also to efficiently manage memory, equipment, and resources. Companies that can optimize these processes will gain additional opportunities to reduce costs.

Against this backdrop, open models can also influence the industry's economics. They expand the range of available solutions and allow developers to adapt technologies to their specific tasks. As a result, competition will depend not only on the capabilities of the models themselves but also on the efficiency of the infrastructure on which they run.

What will determine the success of companies in the AI market

The development of artificial intelligence depends on companies' ability to adapt to rapidly changing technologies. At the same time, investments in physical infrastructure require long-term planning: data centers and energy facilities cannot be deployed at the same speed as software solutions are updated.

Therefore, market participants must simultaneously consider current demand, future model capabilities, and the payback period for infrastructure.

In the long term, competitive advantage will go to companies that can effectively combine energy resources, computing equipment, and software technologies. The coordinated operation of these elements becomes one of the key conditions for scaling artificial intelligence.