OpenAI and Anthropic: The Stock Market Illusion of AI at $852 Billion

AI and the Stock Market Promise: A Echo of Bitcoin in 2010

The rise of artificial intelligence strangely resembles the ascent of Bitcoin in 2010. At that time, a modest investment in Bitcoin could turn an individual into a millionaire within a few years. Today, with the announcement of the IPOs of OpenAI and Anthropic, a new opportunity seems to be emerging. These companies, which have radically transformed the technological landscape, are attracting attention with their mind-boggling valuations. However, upon closer examination of the numbers, one might wonder if the current enthusiasm is not based on a subtle manipulation of perceptions.

When OpenAI and Anthropic announced their intention to go public, it triggered an instinctive reaction in many, reminiscent of the time when Bitcoin was seen as a unique opportunity. Back then, it was enough to invest a few hundred euros to hope to become a millionaire years later. Those who missed that chance still speak of it with a mix of regret and disbelief. Today, history seems to be repeating itself with these two companies that have redefined the world of computing.

A Timely Shift in Discourse

On May 26, 2026, a turning point occurs. Sam Altman, the CEO of OpenAI, publicly admits to being wrong about the impact of AI on employment, stating that the job apocalypse will not happen. Jensen Huang of Nvidia also criticizes companies that use AI as an excuse to lay off workers. Shortly after, OpenAI files its S-1 with the SEC, boasting a valuation of $852 billion. This shift in discourse appears perfectly timed with the preparations for the IPO, aiming to reassure regulators, the public, and investors.

During a conference in Sydney, Sam Altman expressed that he was "delighted" to have been wrong about the apocalyptic forecasts regarding employment. On the same day, Jensen Huang, the head of Nvidia, labeled the rhetoric of leaders who lay off workers citing AI as "lazy." Two weeks later, OpenAI submitted its S-1 to the SEC, revealing a valuation of $852 billion.

Fear as an Economic Driver

Since the emergence of ChatGPT, leaders have perpetuated the fear of massive job destruction, claiming that AI could transform 90% of jobs. This fear has driven millions of workers to invest in AI training, while companies rushed to adopt costly AI solutions. Training budgets have skyrocketed, fueling a thriving market for AI service providers. In reality, this fear was a carefully orchestrated product, with the public as its target.

Dario Amodei, a rival of Sam Altman at Anthropic, spoke of an AI capable of transforming 90% of jobs, further fueling this carefully maintained fear. This fear led millions of employees to invest in AI training, while companies subscribed to expensive licenses to avoid falling behind.

Strategic Turnaround Before the IPO

As the IPO approaches, the discourse changes radically. Sam Altman takes pride in having been wrong about the impact of AI on employment, a reversal that seems calculated to ease fears and attract investors. Meanwhile, the numbers reveal a different reality: in May 2026, 115,000 jobs were cut in the tech sector, with companies like Meta and Microsoft citing AI as the main reason. The contrast between the soothing rhetoric and actual job cuts raises questions about the motivations behind these statements.

On May 26, 2026, Altman stated that he was "delighted" to have been wrong about the impact of AI on employment. However, in May 2026, 115,000 positions were eliminated in tech, with Meta laying off 8,000 people, Microsoft 8,750, and Cisco 4,000. Standard Chartered also announced several thousand layoffs due to AI.

Stratospheric Valuations Despite Colossal Losses

OpenAI boasts a valuation of $852 billion, despite projected annual losses of $14 billion. Profitability is not expected until 2029, or even 2030. In comparison, when they went public, Meta and Uber had much lower valuations and were close to profitability. Anthropic, with a valuation of $965 billion, even surpasses OpenAI, highlighting a frantic race towards ever-higher valuations, with no guarantee of short-term profitability.

OpenAI is valued at $852 billion, despite projected losses of $14 billion in 2026. Profitability is expected by 2029 at best, or 2030 according to the most cautious analysts. The price-to-sales ratio is 65 times the 2025 revenue.

A Closed and Self-Sustaining Ecosystem

Massive investments in AI, estimated at $500 billion in 2026, fuel an ecosystem where AI finances its own infrastructure. The funds invested are used to purchase GPUs from Nvidia and computing power from Microsoft, Amazon, and Google. This closed ecosystem is self-sustaining, with each dollar invested reinforcing the computing capacity that justifies the valuation, thus attracting new investments. The crucial question is who truly benefits from this dynamic.

The $500 billion invested in AI in 2026 primarily goes to Nvidia for GPUs, and to Microsoft, Amazon, and Google for computing power. This closed ecosystem self-values each dollar invested, justifying the valuation and attracting new investments.

Exaggerated Promises of New AI Models

The latest AI models, although improved, continue to do what they already did, but more efficiently. The advancements are undeniable, but they do not represent the often-touted anthropological revolution. The discourse surrounding AI, which promises a civilizational transformation, primarily serves to justify high valuations that do not rest on solid financial fundamentals.

The developments in AI models are real, with notable improvements in reasoning and coding. However, this is not the announced anthropological revolution, but rather an industrial optimization. The discourse around AI serves to justify valuations that do not hold up against an analysis of financial fundamentals.

A Speculative Bubble Like 2000

OpenAI's valuation rests on a seductive narrative of imminent superintelligence and exponential growth. This situation echoes the internet bubble of 2000, where companies like Pets.com reached astronomical valuations without a viable business model. The difference with OpenAI lies in the reality of the technology, but the valuation itself remains debatable, especially in the face of giants like Google and Microsoft that can compete with their own resources.

How can a company that loses $14 billion a year be worth $852 billion? By selling a story of imminent superintelligence and exponential growth. It’s the same mechanism as the internet bubble of 2000, with valuations based on promises rather than solid fundamentals.

Lessons from a Hummingbird Tale

The story of the hummingbird, which tries to extinguish a fire with drops of water, illustrates the importance of critical thinking in the face of dominant narratives. By accepting the narratives of tech companies without questioning them, we risk ceding power to AI. Every individual who chooses to look at the real numbers and ask critical questions helps preserve an essential part of discernment. Technology is powerful and useful, but it should not be accepted blindly at the expense of rational analysis.

My 8-year-old daughter told me the story of the hummingbird trying to extinguish a fire with drops of water. This story illustrates the importance of not abdicating our critical thinking in the face of tech companies' narratives. Every individual who questions the numbers and refuses to yield to the dominant discourse contributes to preserving our discernment in the face of AI.