AI Bubble vs. Dot-com Bubble: Similarities and Differences

A technological déjà-vu? In recent months, discussions about a possible "AI bubble" have become increasingly intense. Many analysts compare the current enthusiasm to the dot-com fever of the '90s, when the Internet promised to change the world. The study conducted by IntuitionLabs shows that, although there are obvious parallels, today's reality is more complex. Market values and profitability In 2000, the Nasdaq-100 index reached a P/E ratio of almost 60x, compared to just 26x in November 2023. Only 14% of listed companies were profitable then, while current AI leaders – NVIDIA, Apple, Microsoft – have consistent profit margins (NVIDIA: 53%). However, many AI start-ups remain unprofitable, attracting record investments of over $73 billion in the first quarter of 2025 alone, representing 58% of the global venture capital total. Adoption and structural differences If in 2000 only half of American households had Internet access, today over 78% of companies use AI technologies. Unlike dot-coms, AI is already integrated into industries, from retail to finance and healthcare. This provides a real economic base, even if expectations sometimes remain unrealistic. The economic challenge of AI According to the Bain & Co. analysis cited in the study, companies will need approximately $2 trillion in annual revenues by 2030 just to cover AI computing costs, but current estimates show a gap of nearly $800 billion. This discrepancy raises serious questions about the sustainability of current investments. Overheating indicators Record market values: NVIDIA surpassed $3.3 trillion in June 2024, reaching $5 trillion on October 29, 2025 – the first company in history to achieve this market capitalization. Extreme concentration: the top three tech companies (NVIDIA, Microsoft, Apple) represent over 40% of the S&P Tech sector. Perceived overvaluation: 54% of fund managers consider AI stocks "in bubble territory." Disproportionate funding: some AI start-ups are valued at $400 million - $1.2 billion per employee. Industry warnings Even AI industry leaders, such as Sam Altman (OpenAI) and Ali Ghodsi (Databricks), have publicly warned of possible overheating. Ghodsi called his own $10 billion funding round the "peak AI bubble," a sign that even the beneficiaries of the boom are aware of the risks. Conclusion: bubble or revolution? The report concludes that we are not experiencing a faithful repeat of 1999. Although some valuations are exaggerated – NVIDIA's record $5 trillion capitalization surpasses any historical precedent – the fundamentals remain solid: widespread adoption, real revenues, and increased productivity suggest a lasting transformation. In other words, AI may go through a market correction, but it will not disappear: it is the beginning of a new technological era, not just a passing trend. Note: The study is based on data up to October 2025. NVIDIA reached the $5 trillion capitalization on October 29, 2025, becoming the first company in the world to achieve this milestone. Full study link: AI Bubble vs. Dot-com Bubble - IntuitionLabs