Sam Altman Warns of AI Investment Bubble
Sam Altman, CEO of OpenAI, has raised concerns about a potential bubble in the artificial intelligence sector, suggesting it could mirror the Dot-Com crash from the early 2000s.
During a dinner with journalists, Altman remarked, “When bubbles occur, clever people get overly excited about the core of the truth.” He noted that, in his view, investors are indeed overly enthusiastic about AI, while also considering it to be one of the most significant developments in recent history.
Altman compared the current excitement over AI funding to the late 1990s tech boom, where many investors were unduly optimistic about internet startups.
From March 2000 to October 2002, the NASDAQ experienced a dramatic decline, losing nearly 80% of its value. This drop was primarily due to many online companies failing to achieve profitability.
A recent report from Torsten Slok, chief economist at Apollo Global Management, argued that the current AI bubble might be even larger than the internet bubble, as the leading S&P companies appear to be overvalued compared to the 1990s.
On a more optimistic note, Ray Wang, a research director at Futurum Group, disagreed with the bubble assessment. He stated that the fundamentals of AI and semiconductor supply chains remain strong, supporting sustained investment in these areas.
However, he did mention that companies centered solely on potential risks might face difficulties with speculative capital.
Some industry experts believe the situation is more complex than simply categorizing it as a bubble. The Chief Technology Officer of Aethir pointed out the real issue may involve supply chain constraints, with excessive capital chasing insufficient projects. This might lead to inflated prices and hype until the necessary infrastructure develops, which he argued is not indicative of a bubble burst but rather market maturation.
There are concerns that the U.S. might already be in an AI bubble, especially following the rapid emergence of Deepseek, a Chinese tech company claiming its AI models were developed at a fraction of the cost compared to U.S. firms.
Deepseek reported spending less than $6 million to train its AI chatbot, a stark contrast to the billions incurred by its American counterparts.
OpenAI is working towards annual recurring revenues of $20 billion, but according to Altman, it’s still not profitable. The launch of CHATGPT-5 faced significant backlash, prompting the company to revert to the previous GPT-4 model.
During the dinner, Altman commented on the term “AGI” (artificial general intelligence), suggesting it has lost some significance. AGI refers to AI capable of performing tasks that typically require human intelligence, which he hoped could be realized in what was once thought to be a “rationally near future.”
OpenAI continues to attract significant investments; in March alone, it raised $40 billion at a valuation of $300 billion, which exceeds previous funding for tech companies. They are currently preparing to sell about $6 billion in shares, now valuing the company around $500 billion.
Altman expressed ambitions for OpenAI, wishing for it to deploy trillions on expanding its data center. He also mentioned interest in acquiring Chrome if the U.S. government considers divesting Google’s web browser due to antitrust laws.
Moreover, Altman confirmed plans to launch a startup that resembles Elon Musk’s Neuralink, while also responding to competition from Musk’s Xai. He noted some firms claim they have found successful avenues, but he found it amusing to highlight their focus on a “Japanese anime sexbot.” Altman emphasized that OpenAI remains focused on developing genuinely useful applications while being mindful of users’ mental health and vulnerabilities.
