As Nvidia joined hands with OpenAI, the developer of "ChatGPT," the "AI bubble" argument has resurfaced. With this partnership, Nvidia will invest in OpenAI, now its largest customer, and with that money OpenAI can increase its use and purchases of Nvidia products, a cycle that critics say is inflating the AI bubble.
Of course, some argue that AI innovation has entered a new phase, but for now the bubble view is gaining more traction. As a result, some tech stocks saw a pullback on the night of the 24th on the New York stock market.
The AI bubble debate could significantly affect Korea's stock market. That is because this issue is a major factor swaying foreign investor sentiment. Moreover, given that the recent rally has been led by the semiconductor sector, this is news investors should watch.
Nvidia said on the 22nd (local time) that it plans to invest up to $100 billion (140 trillion won) in OpenAI. Their goal is to build a 10-gigawatt (GW) AI data center that can train and deploy OpenAI models using Nvidia's advanced AI chips. Ten GW is equivalent to 10 nuclear power plants.
The data center is expected to require about 4 million to 5 million Nvidia graphics processing units (GPUs). That is equal to the total number of GPUs Nvidia will ship this year and is also twice last year's shipments.
Regarding the partnership, overseas securities circles voiced both concern about an "AI bubble" and the view that the "AI revolution has entered a new growth phase."
According to Bloomberg, some Wall Street analysts assessed the transaction as a form of "vendor financing." That is because when Nvidia provides funds to OpenAI, OpenAI uses those funds to buy Nvidia chips. However, OpenAI is said to be renting Nvidia products instead of purchasing them outright.
Stacy Rasgon, an analyst at Bernstein Research, said, "This partnership will clearly stoke 'circularity' concerns." Jay Goldberg, an analyst at Seaport Global Securities, also said the transaction smacks of circular financing and could symbolize "bubble-like behavior."
By contrast, Wedbush Securities in the United States judged that the partnership means "the AI revolution has entered the next growth phase." It said that not only are large-scale capital expenditures expanding at multiple big tech companies, but enterprise AI use cases are exploding across various industries, creating numerous AI beneficiaries across the technology sector.
Dan Ives, a Wedbush analyst, said, "The next three to six months will be when the second-, third-, and fourth-order effects of the AI revolution kick into full gear," adding, "There are concerns about an AI bubble and excessive valuations, but this is like 1996, the early phase of a full-fledged growth cycle for the tech sector." He said it is different from 1999, when the dot-com bubble peaked.
He also said, "While today's AI market is led by a handful of U.S. big tech companies, corporations and governments around the world have joined the AI investment race," predicting that it will expand "from a big tech-led market to Sovereign AI (sovereign AI: AI controlled and operated by a country)."
It is up to investors to decide which outlook to back when allocating assets.