[The opinions expressed in this article are those of the author. They do not purport to reflect the opinions or views of The South Asian Times]
By Vipul Tamhane
In the shiny offices of Silicon Valley, it's a huge situation that is performing a very sneaky financial parlor trick, which might make the 2008 housing crisis look like children's play of accounting. While the globe is fully amazed by the perfect written style of ChatGPT and the whole AI revolution as such, to the greatest extent hidden to the market, the AI money circus story is the one that is getting more and more obvious: these are just circular flows of money, puffing the valuations, and a bubble which is so obscure that it might pull the whole world economy down to new lows.
The details of this calamity are quite simple, but at the same time, they are very surprising due to their courage. Let's imagine a scenario - Nvidia (the company most often linked with the AI revolution) was to harness the power of AI startups such as OpenAI by putting a hundred million dollars into them. The case is, however, not that simple. These investments come with conditions. AI companies should use most of the money received from the investors to buy the high-end Nvidia GPU chips. If Bernie Madoff were alive, he'd be extremely impressed with this financial alchemy.
This is not capitalism. It's a brilliant example of fictional accounting techniques. Nvidia portrays the sales of the chips as revenue, inflating the company's profits and its stock price reaching extraordinary levels. On the other hand, AI companies boast their fundraising rounds as a success of their business models, with a justification for the multi-billion-dollar valuation of their companies, while they continue to lose huge amounts of cash. The emperor has no clothes, but Wall Street is too busy applauding to notice.
One of the most outspoken critics of this is actually Sam Altman, CEO of OpenAI, who, perhaps unintentionally, has exposed this farce to the world. In fact, he said the very astonishing thing that even customers who pay $200 a month, which is a very high sum by any standard, are still unprofitable for the company. The cost of giving the AI services is higher than the income generated. So to speak, the most outstanding example of the AI revolution is the one that loses money on every premium customer. When he was asked about the sustainability of the model, his answer was very direct and surprising: yes, it is a bubble, but the idea is to raise capital faster than your competitors until you are the only one left standing.
Such a confession ought to have resulted in financial markets trembling with horror. Instead, it was received with indifference and the continued energy of the market. The motive behind the act is well known to people who study financial manias, the belief "this time is different", in other words, that AI eventually will transform everything and therefore justify any valuation, loss, or risk.
However, the geo-economic risks arising from this are even scarier. There is no AI bubble just in the boardrooms of California or on the trading floors of Manhattan. Its tentacles are spread in the worldwide financial system, causing risks that far exceed the immediate participants. Tokyo pension funds, Middle Eastern sovereign wealth funds, and small investors from Mumbai to Mexico City have invested in US tech stocks, hoping for the next big thing.
The American economy, which is still the world's largest, has become alarmingly dependent on the AI profitability fiction. The tech sector's disproportionate contribution to US market capitalization means that the next time (not if) the bubble will be breaking, the spread will be quick and brutal. A few AI-related companies have largely driven the S&P 500's recent rally. Remove these overvalued stocks and the reality would be far from being bright.
The high stakes are present equally for the rest of the world. First of all, emerging markets will have hard times if they peg their currencies and set their economic policies in accordance with the dollar without considering the risks of dollar instability. Central banks, on the other hand, will be shocked to see their reserves turn to dust if they have invested them in the US stock market. The global financial system, which depends on the assumption of American market stability, is about to undergo a severe crisis.
The ironically harsh aspect of the situation is that AI is not the cause of the problem. On the contrary, the technology is legitimate, revolutionary, and, eventually, quite helpful. AI-powered tools are now used by millions of people worldwide for work, creativity, and problem-solving. However, the terms utility and adoption do not necessarily imply profitability or valuation. Just to mention one example, the railroad revolution in the 19th century changed society, but most of the railroad companies went bankrupt. The dotcom boom gave birth to Amazon and Google and destroyed thousands of ventures and trillions of wealth on the way.
This bubble's characteristic that makes it particularly evil is the intentionally obscure structure. The circular flow of capital between chip manufacturers and AI companies hides the fundamental aspects of economics. To what extent are the record revenues announced by Nvidia a result of real demand, and to what extent is it due to the reinvestment of capital? When AI startups disclose fresh funding rounds, how much of it is a business model trust and how much is speculative frenzy fear of missing out?
The calculations behind the current log of affairs point to an eventual collapse. If OpenAI loses money at a rate of $200 per user monthly, what will happen when the free offerings are no longer available, and the public will be asked to pay? Will students keep subscribing? Will small businesses be able to afford the cost? Will users from emerging markets, where $200 is a significant part of the monthly income, choose to participate at all? The answer is clear; nevertheless, markets continue to anticipate hockey stick growth.
The way forward calls for an honest and uncomfortable conversation involving all stakeholders. Regulators have to compel the AI industry to disclose the real economics of AI services and corporate cross-investments. Investors should not forego valuation discipline measures; instead, they should suspend disbelief only to a limited extent. The media should question the hype with a skeptical attitude instead of trusting and supporting it without any reservation. And finally, the global financial system needs to get ready for a correction that can alter the markets for a long time.
The query whether AI is going to revolutionize the world is already moot, as the answer is yes. The real question is whether we will let this fundamentally sound technology be utilized as a weapon, leading to a financial crisis by the same speculative forces that have caused market collapses repeatedly. Silicon Valley is overspending, and the rest of the world will need to write the check. The time comes when the world cannot avoid the impact, and it needs to get ready for it.
------------------------------------------------------
.jpg)
(Vipul Tamhane is a counter-terrorism expert and governance consultant)
