원문 보기
This table illustrates it perfectly. Large companies have extended the estimated lifespan of their hardware in their financial reports. And what is achieved by this? Are we facing a true technological revolution or, on the contrary, the biggest speculative bubble of our era? Let's look at the data to try to understand it.
The truth is that if you ask around about investment in AI, the word you hear most is that one, bubble. This is the general sentiment among both industry experts and the public. And be warned, this comparison is no joke. Some analysts are already calling it the biggest bubble in the history of technology investment, which gives us an idea of the scale of speculation we 're talking about.
Furthermore, there is a widespread feeling that something doesn't quite add up . The perception is that everything is sustained by a game in which executives pass the same money from one to another to make the figures appear much better than they actually are. And this is where things get interesting.
If everyone seems so clear that this is a bubble, why do the most powerful and informed investors on the planet continue to pour billions into it every month? What do they see that escapes the rest of us? Well, to begin untangling this mess, the first thing to do is follow the money trail. We need to see where it comes from, because unlike other bubbles in the past, this one has a very, very different financial basis.
The fact is that over the last two decades, tech giants have accumulated a colossal amount of cash . We're talking about companies like Microsoft, Meta, Apple, or Alphabet that have more liquidity than almost any other company in the world. Let's put it into perspective. The dot-com bubble inflated with debt and money from small investors who bet on companies that in many cases did not generate even €1 in revenue.
Today, however, the AI boom is being financed with the money that the technology companies themselves already have. This creates a much stronger tax base . But of course, just because the foundations are solid doesn't mean the building constructed on top will be stable. And this is where we see the most curious thing, how that money moves.
It's a rather peculiar form of financial engineering. A somewhat worrying trend has emerged known as circular operations. It's nothing new. In the days of dining rooms, but now it has been given a much more elegant name, strategic investment. This is the key. Large companies invest in their own customers so that those customers will then buy from them.
For example, Nvidia invests in AI startups, and those startups use the money to buy chips from Nvidia. Google and Amazon invest in Anthropic, and Anthropic charges for using its cloud. On paper everyone's income is growing, but in reality the money is just going around in circles. Now , let's be fair.
There's another way to look at it. This strategy, which may seem suspicious, could also be a very clever way to reduce risks. By investing across the entire chain, they ensure they get a piece of the pie if AI ultimately succeeds. It would be as if AOL had invested in Amazon back in 1999. But the movement of money is only one piece of the puzzle.
Accounting is also playing a key role and could be inflating the apparent financial health of the entire industry. The scale of the investment is staggering. It is estimated that by 2025 alone, large technology companies will spend around $400 billion on the infrastructure needed to train and operate all these AI models.
And if we look at the long term, the figures are almost like science fiction. It is estimated that building all the necessary AI infrastructure could cost around $7 trillion. To give you an idea, this renders historical projects like the Polo program meaningless. And here we arrive at a textbook accounting contradiction.
On the one hand, Nvidia tells us that it will release a new chip every year, which makes the old ones obsolete very quickly. But on the other hand, cloud companies are reflecting just the opposite in their accounting books. This table illustrates it perfectly. Large companies have extended the estimated lifespan of their hardware in their financial reports.
And what is achieved by this? It's very simple: annual depreciation costs are reduced, and as if by magic, the profits seem much higher than they actually are. Very good. After so much financial talk, the question is, how does all this translate into the real world? How does this affect us? Well, the consequences are very tangible.
Market concentration is extreme. A handful of technology companies, the famous "magnificent seven," already account for nearly 34% of the entire SAMPI 500 index. This means that the pension funds of many people are heavily dependent on them, creating a systemic risk for the entire economy. If this sector decides to falter.
We are already seeing the impacts. Energy consumption in data centers has skyrocketed. There are entire communities being surrounded by these mega-constructions. And perhaps most revealing, an MIT report uncovered something incredible. 95% of generative AI pilot projects in companies are not generating measurable benefits, and this leads us to the gap that exists between what is promised and what is actually happening.
For many complex and truly critical tasks, the AI technology we have today simply falls short. This raises doubts about whether such a massive investment is truly justified by their current capabilities. So we return to the central question. Are we experiencing a boom financed with real money from tech companies?
Yes, but sustained by questionable financial engineering, creative accounting, and business results that are currently disappointing. The question that remains is whether we are laying the foundations for a new economy or whether, on the contrary, all of this is a very, very expensive illusion.
Comments 0