Massive investment in artificial intelligence is driving tech sector valuations to historic levels, but insufficient revenue raises concerns about a potential bubble that could severely impact the U.S. economy. Major AI companies insist their technology will revolutionize society, comparing it to milestones like the railroad or the internet. However, these analogies also remind us that such innovations led to financial bubbles before stabilizing. Today, several experts warn that the tech industry could trigger a crisis of significant magnitude, similar to the Great Depression. The rise of AI is not only responding to historical patterns. In the U.S., half of the projected GDP growth for 2025—around 3%—comes from investments in data centers, chips, and energy, essential infrastructure for AI. Seven tech giants, led by Nvidia and Alphabet, account for 35% of the S&P 500's value, a concentration higher than during the dot-com bubble. This mutual dependence between the U.S. economy and big tech is a double-edged sword. On one hand, investment in AI stimulates growth and wards off recession. On the other, if expectations are not met, the impact could drag down the entire economy. Silicon Valley has invested between $300 billion and $400 billion in AI this year, surpassing the GDP of most countries. This figure is expected to reach $700 billion by 2026. Sam Altman of OpenAI aims for a computing capacity of 250 gigawatts by 2033, similar to India's electricity consumption. Despite these investments, sector revenues are far from compensating them. According to Bain, to balance the scales, the industry would need to generate $2 trillion annually by 2030, more than the combined revenues of Amazon, Apple, Google, Microsoft, Meta, and Nvidia in 2024. Journalist Karen Hao states that the current model is unsustainable. The fear of a bubble is based on the possibility that corporate demand for AI may be lower than expected, making profit projections overly optimistic. Additionally, much of the investment is funneled through circular deals: companies like Nvidia invest in AI startups that, in turn, buy their chips. Big tech companies can withstand miscalculations due to their liquidity, but some, like Google, Meta, and Oracle, are already resorting to debt for financing. For startups like OpenAI or Anthropic, which are not yet profitable, pressure will increase if they go public in 2026. OpenAI exemplifies the sector's uncertainty. Three years after the launch of ChatGPT, it leads the chatbot market and is valued at $500 billion, but it remains in the red. Its investment in data centers and chips exceeds $1.3 trillion, and it does not expect profits until 2030. By 2028, its operating losses could exceed $74 billion, according to The Wall Street Journal. The sector justifies its spending with two promises: that AI will boost productivity and lead to superintelligence. However, both are in question. A study from MIT revealed that 95% of AI pilot projects in companies have little or no impact on their results. Moreover, experts like Daron Acemoglu and Ramón López de Mántaras consider the arrival of superintelligence unlikely in the short term. Despite this, these promises attract investments. Big Tech valuations are based on the expectation that current spending will generate future profits, which has driven their stock prices to unprecedented levels. Silicon Valley continues to bet on the future. While the idea of an AI bubble gains traction, not everyone agrees that the sector is overvalued. Sundar Pichai of Google acknowledges some "irrationality" in investment and warns that no company is safe from a crisis. Sam Altman of OpenAI admits that investors are "too enthusiastic" and that someone will lose a lot of money. Wall Street avoids labeling the situation as a bubble to maintain optimism, but markets show signs of nervousness. CoreWeave, an AI cloud service provider, lost $33 billion in value in six weeks, and its shares fell more than 56% in the second half of 2025. Oracle also suffered a 40% drop after going into debt to invest in AI, although both ended the year positively. History shows that bubbles are inherent to disruptive technologies. Carl Benedikt Frey of the Oxford Internet Institute argues that the bursting of a bubble can drive efficiency and leave the most useful technologies, as happened after the dot-com crisis, which solidified Amazon. However, these improvements can take years and cause social suffering. If OpenAI becomes a "too big to fail" company, it may require a public bailout, like banks did in 2008. "Bubbles burst, markets fall, investors lose, and people suffer," warns Natasha Sarin, a Yale professor.