Sam Altman and the AI Paradox: "Bubble for Others, Trillions for Us."

Artificial intelligence is going through its most critical turning point since the launch of ChatGPT. As the industry works through what analysts call "The Great AI Recalibration," OpenAI CEO Sam Altman has made statements that seem contradictory but reveal a precise strategy: warning of an AI bubble while announcing trillion-dollar investments for his own company.

The CEO Who Cries "Wolf" While Investing Trillions

At a dinner with journalists in August 2025, Altman stated bluntly: "Are we in a phase where investors as a whole are overexcited about AI? My opinion is yes. Is AI the most important thing to happen in a very long time? My opinion is also yes".

But here's the twist: In the same conversation, Altman announced that "you should expect OpenAI to spend trillions of dollars building data centers in the not-so-distant future."

As Fortune noted with a certain irony: "What could be frothier than proposing a multi-trillion-dollar expansion in an industry you just called a bubble?"

Two months later, this strategy was confirmed even more clearly. At DevDay on October 6, 2025, Altman stated that profitability and revenue generation "are not in my top 10 concerns" at the moment.

He announced that ChatGPT has reached 800 million weekly active users (up from 700 million in August), that OpenAI is now valued at 500 billion dollars – the highest valuation ever reached by a private company – and that the company is "in an investment and growth space."

The Strategy of "Calculated Misunderstanding"

There is one revealing detail that demonstrates Altman's strategic awareness: he repeated the word "bubble" three times in 15 seconds, joking that the comments would probably become a headline.

It is not naiveté. He is a CEO who overcame an attempt by the OpenAI board to fire him in November 2023, calling it "a major governance failure by well-meaning people, including me," but making him "a more thoughtful leader." Altman knows exactly how the media works and the impact of his words on the markets.

The strategy is clear: let the misconception that fuels the "even the CEO of OpenAI says it's a bubble" narrative run its course, while his company continues to raise billions and consolidate market leadership.

The Sora 2 Case: This strategy emerged even more clearly with the launch of the Sora 2 app in late September 2025. OpenAI launched a TikTok-style social app that initially allowed users to generate videos of copyrighted characters (Mario, Pokémon, anime characters) without restrictions. After the predictable tsunami of controversial content – including deepfake videos of Altman himself eating Pikachu – OpenAI quickly backtracked, switching from an "opt-out" system to an "opt-in" one for rights holders. It's the quintessence of "moving fast" while preaching caution.

Bezos Enters Debate: "Industrial Bubble" vs. "Financial Bubble"

On October 3, 2025, at Italian Tech Week in Turin, Jeff Bezos added a crucial perspective to the debate, confirming the existence of an AI bubble but providing a fundamental distinction that Altman had not made explicit.

Bezos's Definition: "This is a kind of industrial bubble, as opposed to financial bubbles," Bezos explained in conversation with John Elkann, chairman of Ferrari and Stellantis. "The ones that are industrial are not nearly as bad. It can even be good, because when the dust settles and you see who are the winners, society benefits from those inventions."

Bezos provided concrete examples: the biotech/pharma bubble of the 1990s made investors lose over 40 billion dollars collectively, but society gained life-saving drugs. Similarly, the dot-com bubble left behind the fiber optic infrastructure that powers the modern internet – even though the companies that built it went bankrupt.

Bezos noted how the excitement over AI is creating an environment where any experiment, any company - both bright ideas and mediocre ones - is being funded. Investors, swept up in the excitement of the moment, are struggling to distinguish between real opportunities and mere speculation.

He then cited a telling example (without naming names): a six-person AI startup receiving billions of dollars in funding with a valuation of 20 billion. The anecdote was not random. Bezos was highlighting one of the most distinctive phenomena of the AI industrial bubble: revenue per employee.

In the traditional software world, $500K in revenue per employee was considered excellent. Some AI startups are now achieving $3-5 million per employee-an order of magnitude higher.

Tiny teams with powerful technologies can generate huge economic impacts, justifying valuations that by last decade's metrics would seem irrational.

"Very unusual behavior," Bezos commented, but it "typifies the current moment."

The Key Differences Between Bezos and Altman

The two views, while agreeing on the existence of a bubble, diverge significantly in tone and substance.

Altman emphasizes the moderation of enthusiasm, the risk of "burn out," the need for caution. Bezos, on the other hand, emphasizes the "gigantic benefits" to society, the enduring infrastructure, the long-term value that will emerge regardless of individual failures.

The historical perspective also changes: Altman uses generic parallels to the dot-com bubble as a warning, while Bezos provides concrete and detailed examples of how industry bubbles have created permanent value-from fiber optics to life-saving drugs.

Perhaps the most significant difference is a strategic one. Altman controls a company that must continue to raise massive capital to sustain its explosive growth. Bezos has already built the Amazon empire and invests in AI from a position of established strength, with less pressure to justify stratospheric valuations in the short term.

David Solomon, CEO of Goldman Sachs, also present at Italian Tech Week, added, "I would not be surprised if we see a drawdown in equity markets in the next 12 to 24 months. I think there will be a lot of capital employed that will not generate returns."

Bubble or Plateau? A Critical Reflection

But is it really a "bubble" in the traditional sense, or are we witnessing something different?

The Evidence of the Plateau

The Gartner Hype Cycle 2025 for artificial intelligence offers an illuminating perspective: GenAI has slipped from the "Peak of Inflated Expectations" into the "Trough of Disillusionment."

But Gartner is clear: this is not a negative signal. It's a "healthy maturation" that marks the shift from hype to industrial reality. The expected path is 2-5 years to reach the "Plateau of Productivity," where concrete benefits become mainstream.

Data Suggest Plateau, Not Collapse

Industry fundamentals tell a different story from catastrophism:

• 68% of small businesses already use AI with real efficiency gains (not staff replacements)
• 66% of CEOs report measurable business benefits
• IDC forecasts 22.3 trillion dollars in economic impact by 2030
• Every dollar invested in AI generates 4.9 dollars in the economy
• 80% of enterprises will have adopted vertical AI by 2026
• AI governance has risen to become the second most important strategic focus for companies

These are not near-collapse speculative bubble numbers, but technology that is steadily entering enterprise production.

The GPT-5 Lesson: The problematic launch of GPT-5 in August 2025 is illuminating. The model fell short of inflated expectations, Altman had to admit mistakes and restore access to "legacy" models like GPT-4o. But this did not cause the industry to collapse – it simply recalibrated expectations toward incremental growth rather than revolutionary leaps.

The Changing Narrative: From AGI to Pragmatism.

An important sign of the change in strategy is the evolution of the discourse on AGI (Artificial General Intelligence). After years of hype, Altman now calls AGI "not a super-useful term" and "very sloppy."

This represents what Fortune calls "a wholesale turn toward pragmatism instead of chasing utopian visions."

Here again, the strategy is clear: reduce unrealistic expectations that could hurt the industry while focusing on more concrete and measurable metrics.

AI Governance: The New Strategic Priority

While there is talk of bubbles, the real emerging trend is the maturation of AI governance:

AI governance rose from ninth place in 2022 to the second most important strategic focus in 2023, continuing into 2025.

Eighty percent of enterprises have 50+ generative AI use cases in the pipeline, but most have only a few in production.

The Impact on Markets (and Narrative Control).

Altman's comments actually affected the markets: the Nasdaq fell 1.2 percent, Nvidia 3.5 percent, and Palantir nearly 10 percent.

The power to move markets with a simple statement demonstrates Altman's strategic influence. It is the privilege of those at the top: they can afford to "temper" their enthusiasm when it suits them, knowing that their company is solid enough not to be affected.

However, Wall Street analysts remain optimistic. Dan Ives of Wedbush argues, "The AI revolution will fuel a tech bull market for at least the next two to three years. This is a 1996 moment, not a 1999 moment."

The Lesson of Corporate Governance

The layoff experience taught Altman "the importance of a board with diverse viewpoints and broad experience in managing complex challenges."

He now better controls the narrative and message, avoiding surprises that could destabilize his position.

The "selective bubble" is also a risk management strategy: cooling excessive enthusiasm that might attract unwanted regulation or scrutiny, while keeping a high focus on your company's fundamentals.

Recent Developments Confirm Strategy.

DevDay 2025 consolidated OpenAI's position in the developer ecosystem with the launch of AgentKit, ChatKit, and App SDK. The conversation between Altman and Jony Ive touched on the AI device they are developing following the 6.5-billion-dollar acquisition of io – a project facing technical challenges but representing OpenAI's hardware expansion ambitions.

What It Means for Companies

Altman's seemingly contradictory strategy and Bezos' historical analysis converge on some practical lessons for companies navigating this moment.

The first is qualitative differentiation: not all AI investments are equal. Industrial bubbles, as Bezos teaches, don't necessarily reward those who build the infrastructure – the companies that laid fiber optic cable during the dot-com era failed – but the lasting infrastructure they create survives and generates value for society. The crucial question for every company isn't just "how much profit will we make?", but "what impact are we having?"

Crucial is the focus on fundamentals: in an environment where valuations can seem disconnected from reality, the lifeline remains measurable ROI. Companies that demonstrate concrete, quantifiable results survive market corrections.

Vertical specialization emerges as a key competitive advantage. Industry-specific AI solutions are dominating over generic approaches, with results 25% superior according to Gartner.

Proactive governance is no longer a bureaucratic cost but a competitive advantage. Well-structured control frameworks become tools for both risk management and value creation.

For market leaders, a strategic dimension is added: control of the narrative. Those in a dominant position can afford to "moderate" enthusiasm when needed, shaping expectations and industry dynamics.

Finally, the long-term perspective: industrial bubbles, as Bezos observes, create permanent value for society even when individual investors lose money in the short term. The AI infrastructure we're building today will remain usable regardless of market fluctuations.

Conclusions: The Strategic Maturity of AI

The seemingly contradictory comments by Altman and Bezos' analysis reveal a deeper truth: AI is not in a traditional bubble destined to burst catastrophically, but is going through a plateau of selective maturation orchestrated by market leaders.

Companies with solid technologies, clear business models and measurable applications will thrive.

As Fast Company notes, "The CEO captured the situation quite well: only the first part of the quote was turned into countless 'even Sam Altman says AI is a bubble' stories."

The truth is more complex and strategic. We are witnessing the evolution of an industry from the "wild west" stage to one of consolidation. Market leaders use their narrative power to strengthen their dominant position. At the same time, as Bezos notes, they create infrastructure and innovations that will benefit society regardless of market fluctuations in the short term.

FAQ

Q: Does Sam Altman really think AI is in a bubble?A: Altman draws a distinction between overvalued startups without fundamentals and companies with real revenue like OpenAI. He repeated "bubble" three times in 15 seconds, aware of the media impact.

Q: What's the difference between Altman's and Bezos' views on the AI bubble?A: Bezos explicitly distinguishes between an "industrial bubble" (positive, leaves lasting infrastructure) and a "financial bubble" (negative, without fundamentals, collapses with no residual value). Altman uses the term "bubble" more generically as a warning, while Bezos emphasizes the "giant" long-term benefits for society.

Q: Is this a marketing strategy?A: The comments seem calibrated to moderate excessive hype that could attract excessive regulation, while maintaining confidence in one's own company.

Q: Is OpenAI really immune from the bubble?A: With over 20 billion in recurring revenue and 800 million weekly users, OpenAI obviously has stronger fundamentals than many AI startups, but a 500 billion valuation still requires sustained growth.

Q: Are we really experiencing a bubble or is it more of a plateau?A: The Gartner Hype Cycle 2025 suggests a "plateau" – GenAI is in the normal maturation phase toward concrete results. The fundamentals (68% SME adoption, measurable ROI, 22.3 trillion in projected impact) indicate solid industrial growth rather than a speculative bubble.

Q: Are Altman's comments a market strategy?A: Absolutely yes. After the experience of being fired by the board, Altman controls the narrative better. Moderating the hype protects against regulatory risks while keeping attention high on his own fundamentals.

Q: Is vertical AI really the future?A: Yes, Gartner predicts that over 80% of enterprises will use vertical AI by 2026, with ROI 25% higher than generic AI.

Q: Which sectors will see the biggest AI investments in 2025?A: Healthcare, finance, manufacturing, and legal services are leading the adoption of specialized AI, with a focus on applications that demonstrate measurable ROI.

Q: How should investors react to Altman's comments?A: Distinguish between the public message and the underlying strategy. Focus on fundamentals: real revenue, sustainable business models, and consolidated market positions.

Q: Is AI governance really that important?A: Yes, it became the second most important strategic focus for companies in 2023 and continues to grow in 2025, becoming a key competitive advantage and a risk control tool.

Main sources: CNBC, Fortune, TechCrunch, VentureBeat, The Verge, McKinsey, Gartner, PwC, Fast Company, Italian Tech Week 2025