In the late 1990s I was The Economist’s business editor. In those dotcom-boom years my colleagues and I were watching as it became increasingly clear that markets were losing touch with reality. Makers of telecom equipment such as Cisco and Lucent lent billions of dollars to the firms that bought their gear. Later, when dotcom revenues fell short, some of them collapsed, saddling the equipment-makers—and many others—with big losses. Some fear that the same is happening with AI and one of the technology’s keystones, Nvidia. This week on The Insider, we’ll be asking if history is repeating itself or whether this time is different—the most dangerous words in finance.

Nvidia’s rise has been meteoric. It has become the world’s most valuable company, worth more than $5trn, thanks to insatiable global demand for its AI chips. Yet some worry about Nvidia’s financial commitments, which are huge and growing fast. Through sizeable deals, Nvidia provides data-centre landlords and AI labs with cash or guarantees so that they can buy its chips. Underpinning its mountain of obligations are two fundamental assumptions: that its hardware will retain its value and that demand for computing power will continue to surge. I’ll be joined tomorrow by Shailesh Chitnis, our global-business correspondent and a one-time chip designer; Henry Curr, our acting business-affairs editor; and Corbin Duncan, one of our correspondents who focuses on AI, to consider how safe Nvidia’s empire is.

First we’ll look at Nvidia’s rapid rise. The numbers are stunning: its market value climbed from $360bn to $5trn in just three years. Its share price is 14 times what it was when ChatGPT caused a sensation in late 2022. I’ll ask my colleagues to tell us about Jensen Huang, the company’s leather-jacketed boss. He is the magician at the heart of the AI boom, having been one of the people to grasp the technology’s implications early. Shailesh, who recently spent time with Mr Huang in Taiwan, the CEO’s birthplace, can tell us what he is like in person. Has his company become a crucial form of soft power for Taiwan?

Then we’ll evaluate Nvidia’s dealmaking and business model. In August the firm agreed to provide a guarantee of up to $105bn for a vast data centre in Ohio that will use lots of its chips. A week earlier it had joined six big Wall Street firms to raise over $500bn to invest in AI infrastructure, again using Nvidia’s equipment. It could underwrite as much as a quarter of the cost of some of the projects involved. Nvidia’s critics say its financial engineering smacks of dotcom-era “vendor financing”. Nvidia maintains that its schemes could not be more different. I want to get a sense from Henry whether the company’s approach to financing is unusual. And how much competition does the company face? Its latest chips no longer have the market to themselves: roughly half Nvidia’s revenue comes courtesy of America’s cloud-computing “hyperscalers”—Amazon, Google, Meta, Oracle and Microsoft—many of which are designing their own silicon. To what extent are its investments defensive?

Finally, we’ll consider how risky this all is. Nvidia is pledging lots of money on the assumption it’ll either make a profit, or never have to pay out on its guarantees. The big question is whether demand for AI justifies this. Is Nvidia inflating a dangerous bubble? The biggest ten public companies championing AI make up 40% of the value of the S&P 500 index; Nvidia alone accounts for 8%. Will the company’s dealmaking accelerate AI development, or will it deepen a stock-market crash? Could that cause a recession? Banks, investors and the world’s biggest companies are all making the same bet as Nvidia. Are they all wrong?

The show will be available to watch from 6pm London time (1pm in New York) on Thursday. We’ll be reviewing the results of this week’s poll during the discussion, as well as answering some of your questions. Please tell us what you want to know by using the Q&A feature on the episode page. And if you have feedback or suggestions for future episodes, please write to me at insider@economist.com.