Questioning the 'Common Sense' of the AI Boom: Who Are the True Winners and Losers Behind the Overheated Valuations?
The momentum of the generative AI boom shows no signs of stopping. Entering 2025, the market capitalization of AI-related stocks has been swelling without a ceiling. However, a question is spreading among investors: 'Can AI valuations be justified?' Focusing on the trends of representative companies like Palantir, Tesla, and Snowflake, we explore the risks and essence hidden within the AI market.
1. Palantir: A Symbol of the AI Bubble, or an Exception?
1-1. True Competitive Advantage Born from '20 Years of Accumulation'
Despite Palantir showing solid growth in its Q3 2025 earnings, its stock price fell. The reason is concern over its 'overheated valuation.' However, analyst Mariana Perez Mora stated the following on the program:
“Palantir has been building the ‘infrastructure that powers AI’ for over 20 years. That foundation is a strength that no other company possesses.”
The company's AI platforms, 'Foundry' and 'AIP,' are increasingly being adopted in both government and commercial sectors, with the U.S. commercial division recording 121% year-over-year growth. The fact that AI implementation is actually bringing about operational efficiency and cost reduction sets it apart from many 'expectation-driven' companies.
1-2. Which Companies Will Survive After the 'AI Bubble Bursts'?
Mora continues by pointing out the following:
“Even if the AI bubble bursts, Palantir will survive. This is because the company is proving not ‘what can be done’ with AI, but ‘how it creates value.’”
As investor Michael Burry places 'put options' betting on a decline in Palantir's stock price, this difference in fundamental strength will become a watershed moment moving forward.
2. Tesla: A 'No' to the Massive Compensation Package—A Warning from Investors
The world's largest sovereign wealth fund, the Government Pension Fund of Norway, voted against the compensation proposal for CEO Elon Musk. While ostensibly 'acknowledging his achievements as a visionary,' it clearly expressed concerns regarding
the excessive scale of the compensation
and 'key-man risk' (dependence on Musk).
According to Bloomberg reporter Craig Trudell, the fund has cast similar dissenting votes in the past, and there is a history of Musk 'retaliating with criticism.' Tesla's stock price fell 4% at one point, pouring cold water on expectations for its lead in the AI and autonomous driving sectors.
3. Snowflake and the Essential Value of 'AI Infrastructure Companies'
3-1. Aiming for a New Layer of 'Cloud x AI'
Snowflake CEO Sridhar Ramaswamy is in the spotlight as a company 'seizing control of the data layer in the AI boom.' Through a partnership with Google Cloud, he emphasized a stance of treating AI models 'neutrally,' such as by making the Gemini model available on the Snowflake platform.
He states:
“We are providing AI not as a ‘spectacle,’ but as a practical tool that allows customers to achieve ROI (return on investment).”
3-2. A Field Perspective on the 'Bubble Debate'
Ramaswamy responds to the question of whether an AI bubble exists as follows.
“The market is certainly enthusiastic. However, the one point we must focus on is whether customers are truly gaining value. External evaluations are nothing more than noise.”
This suggests that a business model backed by actual demand is the condition for survival after a bubble bursts.
4. Investor Sentiment and Macro Risks: The 'Second Chapter' of the AI Bubble
Kristina Hooper, Chief Global Market Strategist at Invesco, says, 'The current state of AI investment is reminiscent of the dot-com era of the late 1990s.'
“A mentality is spreading that anything AI-related will go up. However, if it is not accompanied by actual effectiveness in productivity improvement or cost reduction, investment fever could cool down all at once.”
Furthermore,
rare earth supply constraints
rising power costs for data centers
and 'NIMBY' (Not In My Backyard) movements (local opposition) are also becoming obstacles to AI expansion.
Conclusion: The Future of Valuations Will Be Determined by 'Implementation Capability'
The debate surrounding the valuation of AI-related stocks is not merely about 'bubble or not.' What is truly being questioned is 'which companies can connect AI to actual profits in the field.'
Companies like Palantir, which are demonstrating value creation using 20 years of accumulated experience as a weapon, and companies like Snowflake, which are organizing data infrastructure and visualizing ROI—these 'implementation-type AI companies' are the ones whose true worth will be tested in the next adjustment phase.
The 'second chapter' of the AI boom will likely be led by companies that grasp reality rather than illusions.
