OPINION:
Stock market gains are fueling wealthy consumers, investments in artificial intelligence and economic growth.
Many consumers, coping with inflation and uneasy about AI at work, are trimming their sails.
The wealthiest 10%, who account for almost half and a growing share of consumer spending, have reaped huge stock market gains, elevating the economy. This year, stock market appreciation among the top 30 AI companies has created $5 trillion in additional wealth.
That’s about half the overall $10 trillion stock market gain, and the “wealth effect” should boost consumer spending by $350 billion to $500 billion.
That’s why we see lines to enter Chanel boutiques on Saturdays at the mall while Macy’s shutters stores.
Uncertainty about President Trump’s tariffs makes businesses reluctant to invest outside the AI space.
Shutdown delays in government data notwithstanding, industry reports indicate computer equipment, software and data center construction for AI are driving the lion’s share of growth in business investment.
However, investors whose sentiments drive stock prices are getting nervous because so many of the gains are concentrated among the “Magnificent Seven”: Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia and Tesla.
In turn, those are plowing much of their profits into speculative ventures such as OpenAI and in-house AI, whose potential for monetization may not justify their scale.
Next year, Alphabet, Microsoft, Amazon and Meta expect their AI investments to exceed $400 billion. Based on all the hype, OpenAI has a market capitalization of about $500 billion. It should have only $13 billion in revenue this year and projects operating losses will swell to $74 billion by 2028.
The deals it has made for cloud and computing capacity commit it to spending approximately $1.4 trillion over the next eight years, raising questions about whether it will ever turn a profit.
OpenAI sells business applications, but a good deal of its resources and those of firms such as Alphabet and Meta focus on achieving artificial general intelligence, machines that can think as well as human beings and much more rapidly.
Imagine Apple’s Siri with the IQ of a Nobel laureate, with specialized editions offering your child Taylor Swift and Albert Einstein as music and physics tutors.
At the center of all this are the remarkable chips designed by Nvidia and its competitors, such as Advanced Micro Devices, which are fabricated by the Taiwan Semiconductor Manufacturing Co.
Nvidia is the world’s most valuable company, but how much of its projected sales are authentic? Recently, it struck a deal to invest $100 billion in OpenAI, which helped Sam Altman build his $1 trillion data center empire.
It will cost at least that much to father a Swift or Einstein and make them as ubiquitous as a Ford F-150. In return, OpenAI is committed to spending a good deal of that sum on Nvidia chips.
This pattern repeats quite a bit in Nvidia’s investments in firms that build data centers and develop specialized AI agents.
Agents perform discrete tasks such as computer coding and legal research and work in teams that communicate with one another. Those are raising productivity in activities as diverse as banking, apparel design and marketing.
Look at what’s happening at Nvidia.
It’s paying OpenAI to purchase its chips. Those have a useful life of perhaps five years, and OpenAI is likely to still not have achieved profitability.
Nvidia gets to book a sale on its income statement and assets in the form of OpenAI stock on its balance sheet. Those shares are worth something only if OpenAI indeed delivers Einstein.
If not, Nvidia has squandered $100 billion on eventually worthless stock.
Less noticed is that OpenAI’s principal competitor, Anthropic, is more commercially minded. Anthropic has larger market shares for coding applications and enterprise sales of large language models than OpenAI, Google and Meta and has avoided OpenAI’s forays into image and video generation, which require more data center capacity.
According to an analysis by The Wall Street Journal, it has a decent chance of turning a profit by 2028.
Palantir is also more focused on applications to help government and business clients integrate and analyze complex datasets to assist in operational decision-making. Its AI platform helps organizations build apps, automate tasks and develop their own specialized AI agents.
In the third quarter, it had $1.18 billion in sales — a 68% year-over-year gain — and $475.6 million in profits with a diversified client base.
I have few doubts that OpenAI will one day be a profitable company, but it may first have to go through bankruptcy reorganization. Otherwise, it could be dismembered with its various intellectual properties sold to firms such as Palantir and software designer Salesforce.
In the end, the economy will gain from AI as much as it has from the internet, and investors should remember that the “Magnificent Seven” have growing profits and can absorb some losses without sinking.
• Peter Morici is an economist and emeritus business professor at the University of Maryland, and a national columnist.
Please read our comment policy before commenting.