The AI Economy’s Fragile Facade

Last week’s stock market turmoil has left investors scrambling to make sense of the opaque and complex world of artificial intelligence (AI). The rollercoaster ride was sparked by a double whammy: China’s memory chipmaker CXMT floated on the Shanghai stock market, soaring 466% in value, while reports emerged that China had developed its own tools for deep-ultraviolet lithography, a technique crucial to the computer chip supply chain. As investors reeled from the news, AI-linked shares plummeted worldwide.

The reliance of the AI economy on Nvidia has become so egregious that it’s starting to resemble a house of cards. The world’s largest listed company is the linchpin holding up vast parts of the global economy and stock market. Its dominance is built on a web of circular and opaque deals, making it increasingly difficult for investors to understand what they’re buying into.

The implications are far-reaching. Chris Beauchamp, chief market analyst at IG, noted that Chinese chip companies are poised to disrupt the big players in the same way they’ve undercut industries like steel and automobiles. But this isn’t just about competition; it’s also about the structural fragility of the AI economy itself. The reliance on Nvidia is a recipe for disaster, waiting to be triggered by even the slightest misstep.

Nvidia’s dealings with OpenAI are shrouded in opacity. A recent report that the company was considering providing a $250 billion backstop to the AI startup sent shockwaves through the market. This comes after a failed deal between the two companies, and investors are increasingly anxious about Nvidia’s central role in the AI ecosystem.

The anxiety is well-founded. Alvin Nguyen, an analyst at Forrester, observed that Nvidia has become a kind of “central bank” for the AI economy, propping up vast parts of the market with its dominance. But this cannot last forever. The gravy train will eventually run out of steam, and investors are waiting with bated breath for the inevitable collapse.

CXMT’s debut may have sent shockwaves through the market, but it’s not a direct threat to Nvidia’s dominance. Manufacturing deep-ultraviolet machines is a “massive symbolic victory” but not a commercial replacement for ASML overnight, according to Mark Boost, CEO of Civo. The real concern lies in China’s ability to develop its own lithography tools, which could potentially disrupt the global supply chain.

However, this is still years away from becoming a reality. Fables take years to develop, and until these Chinese tools can match western reliability, ASML’s dominance remains “structurally safe outside mainland China.” The short-term threat, therefore, is more of an overreaction than a genuine concern.

Long term, however, this week’s advances are a gamechanger for the AI economy. Chinese chip companies will undercut and outcompete their western counterparts on price, as Chris Beauchamp observed. This shift is not just about China; it’s also about the fundamental change happening within the global supply chain.

US export controls have forced China to develop domestic capabilities, but this has been a predictable outcome. The real question is what this means for Nvidia – and the world of AI as a whole. As Nguyen astutely observed, “Everybody’s waiting for them to fall apart.” And when that happens, it will be catastrophic.

The investor skittishness surrounding Nvidia is not just about the company itself; it’s also about the opaque and circular deals that underpin its dominance. The $250 billion backstop being considered by Nvidia for OpenAI is a case in point – and investors are rightfully wary of this kind of arrangement.

In the end, last week’s stock market turmoil has exposed the fragility of the AI economy’s facade. While CXMT’s debut may have sent shockwaves through the market, it’s not the only story here. The real issue lies in Nvidia’s reliance on opaque deals and its role as a linchpin holding up vast parts of the global economy.

As investors wait with bated breath for the inevitable collapse, one thing is clear: the AI economy cannot continue to rely on a single company forever. The house of cards will eventually come tumbling down – and when it does, the consequences will be catastrophic.