This “Is the AI Investment Boom a Financial Bubble?” article was written by Philipp Taubenblatt – Financial Analyst at I Know First.

Highlights

• How cheap money and corporate FOMO inflated the AI boom, pushing companies to chase “AI stories” instead of real value.
• How adoption skyrocketed worldwide, with AI moving from a niche tool to something nearly every company claims to use.
• Why this creates a potential bubble, as hype, expectations, and spending outpace actual productivity gains.
  

Innovation or Irrational Exuberance?

The most significant investment theme over the last decade is, without doubt, artificial intelligence. Nvidia’s market capitalization recently eclipsed $5 trillion, a first. Much of this growth stems from Nvidia’s dominance in AI computing. Since the launch of ChatGPT in late 2022, Nvidia’s stock has risen over 1,000%, with emphasis on data-center chips, which are used in nearly every “major” model. Even in early 2025, AI consumes half of global venture funding, around $70 billion, almost all of which has gone to startups with little or no revenue. OpenAI, which has endured somewhat modest profits, is investigating a $1 trillion IPO by 2026. And, though AI has the ability to change the world, now or later, these types of numbers evoke a similar question to questions raised in past bubbles, are these investors again pricing into today’s stock prices decades of future profits?

The Lifecycle of a Bubble

Minsky, an economist, uniquely characterized speculative bubbles in five often-recurring sections:

Displacement: Excitement and new flows of investment are generated by a new idea or a shift of some sort — a change such as the internet in the 1990s, a housing shift in the 2000s, or AI at present.
Boom: prices begin moving up with some level of consistency as hope increases, the media celebrates, and more money comes rushing in.

Euphoria: Valuations increase, and have a disconnect with any fundamentals. Any naysayers are ignored; “this time it’s different” t-shirt slogans begin circulating.
Profit-Taking: Early investors take some money off the table as they sense this is not going to grow sustainably or forever.

Panic: Confidence evaporates, small disappointments cascade, and the market begins to shed excess.
Historically, these cycles last 5-8 years, meaning that in true Minsky fashion, the AI boom that got started in earnest in 2023 is somewhere between boom and euphoria with an uptick in valuations beyond revenues, and a rush of investor confidence that approaches inevitability, which is typically when bubbles reach their peak.

Cheap Money, FOMO, and the ROI Conundrum

For years, interest rates were almost zero and even the pandemic stimulus pumped cash into the market. The combination of enormous tech companies with their large rates of liquidity and other venture funds was chasing exponential growth stories around AI. It was an easy sell, difficult to define, and had an unlimited story to tell. The continued rate hikes could not slow the momentum. The story that AI is the destiny for eager executives replaced any discipline towards valuation. As other companies began to announce their own AI initiatives out of fear of being left behind, we seen the corporate fear of missing out.

The Shiller P/E compares a stock’s current price to its 10-year average of inflation-adjusted earnings. This method smooths out economic cycles to provide a more stable, long-term measure of valuation, indicating whether a market is overvalued or undervalued by comparing the current ratio to its historical average. Today, the Shiller P/E ratio is rising to levels comparable to those seen during the dot-com bubble of 1999-2000. One reason for this elevated P/E ratio is the expectation that the adoption of artificial intelligence technologies will lead to increased corporate productivity and profitability in the future..

The tech industry’s size rapidly rose at a rate that markets had never experienced during the dot-com era. From 1998 to 2000, tech fair share as a percentage of the S&P 500 doubled, rising from 23% to almost 47% in less than two years, an extraordinary concentration that contributed to the 2000 market collapse. At the high point, many of these companies had forward P/Es above 60-70x with little or no profitability. The current AI cycle is different in circumstances, but similar in structure. By the years 2024-2025, companies with exposure to AI will comprise approximately 44% of the S&P 500, and the prominent firms with AI exposure are trading at forward P/Es around 26-31x. While none of this comes close to the extremes of the year 2000, the same pattern can be seen: rapid expansion in the sector, elevated valuations and increased investments from the  investors, before eventually having any sustainable cash flow.

The application of AI has increased considerably over the past ten years. In 2017, only about one in five companies reported using AI in their products or internal operations, but by 2024-2025, that number had climbed to about 70%-80%. In short, this means that hundreds of millions of companies worldwide have begun integrating at least one AI-enabled process into their operations. And similar to this growth in the number of AI-adopting companies, we have witnessed a boom in scientific output as well. Ever since the mid-1980s, when there were around 4000 AI publications annually, we are now seeing more than 50,000 publications every year reporting on various AI-related topics.

The increase in AI projects does not imply that these projects are successful. Studies indicate that approximately 40%-45% of companies report that most of their AI activities do not even move past the pilot stage. Approximately half of proof-of-concept attempts do not make it to production within the company. Thus, while adoption continues to increase, successful integration and implementation are still bottlenecks.

The fear of missing out on the next wave of the revolution pushed some of even the most cautious investors into the most high-risk companies. Hedge funds and sovereign wealth funds and corporates brought in massive amounts of foam, further building an illusion of profitability from AI in the near future. The buying based merely on the fact that we all wanted to bet the same way created extreme optimism and inflated valuations far beyond any defined fundamentals. It felt somewhat like the bubbles before, like the housing boom of 2006 and the crypto boom of 2021.

But every bubble starts with a base of truth. AI will permeate almost every industry, large and small, except that not every company will build a profitable business around AI. Training increasingly large models costs billions in chips, power and data, and takes time to monetized. AI will remain free or experimental for a long time. Extreme amount of free spending with little if no return will add to the overheated perceptions, we are ultimately too investing into something that exceeds the amount of the return.

Resonances of the Dot-Com Boom

The late 1990s internet boom provides a nearly perfect parallel.

At that time, telecoms had overbuilt fiber before anyone used it. Now, hyperscalers and others are investing their billions in GPUs and infrastructure, long before they can confirm revenue. OpenAI’s CEO, Sam Altman, has publicly stated they need “trillions” of investments in AI. The mindset is the same: scale first, revenue/profits second.

Insiders understand. Bret Taylor, the chairman of OpenAI, said, “AI will change the economy and yes, we are in a bubble. A lot of money will be lost!” The technology is real. The valuations are not. Again, a handful of companies may emerge stronger, while most companies eventually will fail.

Who Survives When the Bubble Bursts

Following the dot-com crash, businesses that had real products, cash flow, and scale—companies like Amazon, Google, and Microsoft—persevered while other companies disappeared. In the same spirit, the companies that will succeed in the AI renaissance will be determined by a market correction. Ultimately, the companies that are likely to survive and thrive are Microsoft, Amazon, Alphabet, and Nvidia: companies with profits, deep moats, and balance-sheets. These companies can acquire badly run and underfunded start-ups and continue to innovate. If funding dries up in this business model, the smaller players will cease to exist.

There is no doubt that AI will be a part of the future; however, the winners will be identifiable only after the dust settles from the current liquidity rush.

An Omen of a Peak

• Optimism: AI writing is prevalent in every sentence from the news to discussions with friends – if you challenge this optimism for AI, you’re simply uneducated.
• Exuberant IPOs: Companies like CoreWeave suddenly go public at $23 billion and then 2X’s the value in months after – pro forma behavior of manic speculative behavior.
• Disconnect From Fundamentals: Startups seeking valuations in the hundreds of millions of dollars per employee, and one investment advisory firm estimated that by 2030, infrastructure spending will have exceeded revenue efficiency in AI by $800 billion.
• Insider Skepticism: When bold and brash figures behind AI enthusiasm like Sam Altman, Ray Dalio, and Baidu’s CEO compare the same wave as 1999, exuberance has peaked.
  These red flags don’t signal the imminent end tomorrow, but they all signal a market that is priced for perfection, and that’s when risk outweighs opportunity at a silent level.

The Outcome: What occurs when it explodes

Every bubble ends in a similar fashion: sudden repricing, a horde of failures, and recoveries that unfold slowly. Once capital becomes constrained, unprofitable companies fail- and as history has shown, bubbles do not destroy technology; they cleanse it. When the weakest actors fail, the stronger players can then acquire talent, infrastructure, and intellectual property for a fraction of the cost, preparing them to emerge as the next decade’s leaders. Excess investment will no doubt sting in the short-run, but it is an investment in future progression. The data centers, chips, and breakthroughs that were built during the excessive investment will lead the next productivity boom once the speculation wears off and the fundamentals reassert themselves.

The Aftermath: What Happens When It Bursts

Bubbles usually have the same outcome that can be put into three words: repricing, failure, and recovery. As capital dries up, projects collapse. The key here is that only unsustainable projects collapse, and as they do, history shows that technology itself does not suffer but it is rather a way of cleansing it.

As failures multiply, consolidation follows. Large companies buy all sorts of valuable IP, infrastructure, and talent at only a fraction of what it was before. Those who come out of that crisis will become the next decade’s most important players.

While it is painful, overinvestment leaves behind long-lasting progress. All the breakthroughs, data centres, and chips that were acquired during that process will be fundamental to the next wave of productivity that will follow. After the crash, the markets rediscover fundamentals. This is where the actual value creation begins.

Seeing Through the Noise: How AI Forecasting Helps

One of the hardest challenges in markets is recognizing when enthusiasm turns to out-of-control exuberance – by the time you have noticed the bubble it’s already popped. The real advantage is figuring out how to identify emotional changes in sentiment before they become noticeable.

AI-based forecasting, like I Know First does, provides that advantage. AI technology has the ability to analyze mountains of historical and current market data to discover when optimism has begun to separate from fundamentals, even before investors recognize it. In an emotional market, driven by fear, greed and collective behavior, one of the best methods for investors to remain rational, and emerge on the other side stronger, is through disciplined forecasting.

Yes, AI is creating excitement and mania in the current cycle, but it is also providing create resources for grasping and ultimately managing it.

Finding Investment Opportunities with AI

I Know First provides stock market forecasts based on chaos theory approaches. Previously, we discussed the Conceptual Framework of Applying ML and AI Models to Analyze and Forecast Financial Assets. The I Know First predictive algorithm is a successful attempt to discover the rules of the market that enable us to make accurate stock market forecasts. Taking advantage of artificial intelligence and machine learning and using insights of chaos theory and self-similarity (the fractals), the algorithmic system is able to predict the behavior of over 13,500 markets. The key principle of the algorithm lays in the fact that a stock’s price is a function of many factors interacting non-linearly. Therefore, it is advantageous to use elements of artificial neural networks and genetic algorithms. How does it work? At first, an analysis of inputs is performed, ranking them according to their significance in predicting the target stock price. Then multiple models are created and tested utilizing 15 years of historical data. Only the best-performing models are kept while the rest are rejected. Models are refined every day, as new data becomes available. As the algorithm is purely empirical and self-learning, there is no human bias in the models and the market forecast system adapts to the new reality every day while still following general historical rules.

I Know First has used algorithmic outputs to provide an investment strategy for institutional investors. Below you can see the investment result of our AI investment strategy for the period from January 1^st, 2020, to July 31^st, 2025.

The strategy provides a positive return of 587.22% which exceeded the S&P 500 return by 491.12%.

Conclusion

Technological revolutions begin with inflated expectations, transition into a bubble, and then stabilize. The market often overestimates the short-term benefits of AI and underestimates its long-term impact. Investors who focus on data and fundamental metrics reap the greatest rewards by distinguishing true innovators from speculators. I Know First artificial intelligence algorithm has proven itself as an effective tool for identifying the most promising investment opportunities in current market conditions.

To subscribe today click here.

Please note-for trading decisions use the most recent forecast.