AI fever and who makes money from it?

Looking at the results of the fiscal year 2025, known for NVIDIA's graphics chip production and an increase of 114% y/y, it can be seen that just as during the gold rush the best paid were shovel suppliers, today in the age of AI fever, computing power suppliers earn the best.

, “The demand for Blackwell is incredible as inferential artificial intelligence (reasoning AI) introduces the next stage of scaling — increasing computing power for training makes models smarter, and increasing computing power for long-term thinking makes answers smarter,” said Jensen Huang, founder and CEO of NVIDIA in an official company report.

“We were able to increase the mass production of Blackwell AI supercomputers, achieving sales of billions of dollars in the first quarter. AI is advancing at the speed of light, and agent and physical AI are setting the stage for the next wave of AI that will revolutionize the largest industries.”

The fact is that Blackwell's performance and packing 208 billion transistors in a single chip is impressive, just as impressive is the technological excellence of the 4NP (developed 5nm technology) of the Taiwanese company TMSC, but in NVIDIA's results you can see what is the scale of investment in the computing power necessary to train increasingly advanced AI models.

This begs the question, who is making money on it? For example, one of the flagship companies offering AI models, OpenAI, in the first half of 2025 had revenues of $4.3 billion, a loss of $7.8bn and a cash burn of $2.5bn.

The scale of the investment so far can be imagined taking into account the time of three years that has elapsed since the launch of the ChatGPT service in November 2022 and how long it took to prepare the site before its launch.

We see interest in the possibilities and the prospect of using AI in practice in many industries, hence our investments in the practical use of the capabilities of artificial intelligence in our low-code platform Meltemee. The application areas of AI models are, on the one hand, support for the application designer and tools that shorten the production cycle, and on the other hand help for the user in communication and AI agents.

The Meltemee platform provides a good foundation for building agent AI systems. Thanks to its flexible infrastructure, a designer without a technical background can create tools for agents as low-code processes. It is possible to automatically share them with artificial intelligence, using the MCP protocol, without complex configuration or writing code.