Three years have passed since the euphoria around ChatGPT began, but skepticism is growing on the exchange: companies are spending more and more on models and data centers, and there is little evidence of a quick payback.
Three years have passed since the euphoria around ChatGPT began, but skepticism is growing on the exchange: companies are spending more and more on models and data centers, and there is little evidence of a quick payback.
According to Bloomberg, investors are nervously reacting to signs of overheating, including Oracle’s plunge after reporting higher-than-expected infrastructure spending. The debate is centered on whether the company will run out of fuel at a critical moment.
One risk is access to capital. The article notes that OpenAI plans to spend $1.4 trillion in the coming years, and The Information estimated a possible «burn» of $115 billion by 2029, with a positive cash flow in 2030. If investors start giving money more cautiously, it will hit both the businesses around OpenAI and the entire infrastructure chain.
The second factor is the scale of Big Tech spending. Bloomberg Intelligence predicts more than $400 billion in capital investments by Alphabet, Microsoft, Amazon and Meta over the next 12 months. The risk is that such spending puts pressure on free cash flow and could limit buybacks and dividends.
At the same time, Bloomberg emphasizes: these are not «dot-coms» yet. The Nasdaq 100 is valued at about 26 times projected earnings, while at the peak of the dot-com boom the figure exceeded 80. However, there are overheated niches: Palantir is trading at more than 180 times expected earnings.
Leonid Polupan, who worked at Microsoft and now at the AI company Seeton, called the market overheated in an interview with Forbes, but not a classic bubble: some investments go into material things, and the geopolitical race between the US and China forces people to continue investing even in the event of a correction.
Previously, dev.ua wrote about how in the third quarter of 2025, Nvidia showed record quarterly revenue of $57 billion and sharply increased its forecast, stating that the artificial intelligence market is in a phase of sustainable growth, not a speculative «bubble.»
