Charles’ Note: I love ChatGPT, and I use it daily. It’s become an invaluable resource for me and helps me organize information vastly quicker than I could on my own.
It costs me a whopping $20 a month. A pittance.
But here’s the thing…
If they raised the price much, I’d drop my subscription.
Keep in mind, I would be willing to pay hundreds or even thousands of dollars per month for it if there was no alternative. I’d consider it worth every penny.
But why would I do that if I can use a competitor for a fraction of the cost?
Remember DeepSeek, the Chinese competitor to ChatGPT that made a big splash earlier this year?
It’s open source and effectively free. Only large enterprise customers pay… and they don’t pay much.
ChatGPT seems to be a little better than DeepSeek. But the key words here are “a little.”
And that marginal improvement comes at a massive cost.
While the official price tag has never been released, ChatGPT’s latest model cost an estimated $2 billion to $5 billion to develop and train.
AI is changing the world.
We NEED it to change the world. As I wrote in yesterday’s Navigator, we can’t meaningfully grow our economy in the decades ahead without it.
But the economics are terrible.
Once the technology is “out there,” it’s easily replicated and remarkably hard to charge for.
And, as my Bill Bonner of Bonner Private Research points out, the actual return on investment for the companies pouring the most into research and development is dubious at best.
Leave it to Bill to see through the hype. He’s been doing it for longer than I’ve been alive, helping his readers cut through the B.S. Take it from here Bill!
Too Much of a Big Nothing
By Bill Bonner, Bonner Private Research
Two companies – Nvidia (NVDA) and Microsoft (MSFT) – each are worth more than $4 trillion. Together, that’s more than India’s and Japan’s combined annual output.
Price is what you pay, as Buffett puts it. Value is what you get. Our question for today: how much value will investors really get from the Magnificent 7?
Our Law of Conservation of Value tells us that prices cannot stray too far or too long from value. And value depends on output. Investors ought to be able to look to a future stream of income and from it earn their money back… and more.
Even in the dot-com bubble in 1999, the top companies were not as valuable or as concentrated as they are today.
Nvidia, Microsoft, Alphabet (GOOG), Apple (AAPL), Meta (META), Tesla (TSLA), and Amazon (AMZN) – together, these companies make up a third of the total U.S. stock market value… an amount roughly equal to China’s GDP.
Part of the appeal of these Mag 7 stocks is that they are widely believed to be taking advantage of AI technology. In the case of Nvidia, of course, that is the central appeal. But the others are investing heavily in AI too.
In 2024 and 2025, Meta, Amazon, Microsoft, Google, and Tesla will put more than half a trillion into AI. The revenue from these investments is expected to be around $35 billion.
Amazon, for example, has invested more than $100 billion, which is thought to generate an extra $5 billion in revenue.
We don’t know how reliable or meaningful these figures are. What we do know is that they aren’t very impressive.
As in the dot-com boom of the late ‘90s, AI is not paying off.
This is an in-put story, with huge investments made in the hope of creating AI-based wealth. But so far, the output doesn’t measure up.
You can go to ChatGPT, for example, and pay for the service. Many people use it occasionally – including us. But few pay for it – also including us.
This would be fine, except that so much investment has gone into AI development that anything less than spectacular results will look like failure.
One estimate, from Goldman Sachs, for example, showed that the Mag 7 would have to produce $600 billion in extra annual revenue to make sense of their investment.
Michael Roberts:
So while the excitement of AI takes the stock market to new heights… a huge investment of money and resources, astronomical payments to AI trainers, and the construction of huge data centers [there]… so far no significant revenue has been generated and there is almost no profit. This is a steroid-friendly version of the dot-com bubble.
The appeal of the dot-com era was the idea that more information would lead to higher GDP growth rates with less need for capital investment. Costly trial-and-error expansion would be replaced by less costly, more precise, knowledge-driven growth, or so it was believed.
It didn’t work out that way. Productivity and growth rates generally softened throughout the 21st century. Capital investment went down. The Internet/Information Revolution did not compensate for the decline; it seems to have made it worse.
The OECD adds detail:
In the last half century, we have filled offices and pockets with increasingly faster computers, but the increase in labor productivity in developed economies has declined from about 2% annually in the 1990s to 0.8% in the last decade.
Even the production per worker of China, which once increased rapidly, has stopped.
Research efficiency has decreased. Today, the average scientist produces less groundbreaking ideas per dollar than his colleagues in the 1960s.
Despite the rise of intangible assets, total investment has generally been weak since the global financial crisis, which has directly worsened the slowdown in labor productivity.
Will that change with AI?
Probably not.
The defining curse of the Information Revolution was too much information. It piled up. It got distorted and misinterpreted. It took time and money to store and sort. And much of it was either false or useless.
Now cometh AI, adding to the too-much-info problem. Already, it generates news and reports that fill our in-boxes and waste our time. And an Israeli company just announced that it can twist and turn (distort) the news in real time.
Which leaves, at least for now, AI and the Mag 7 in an old-fashioned financial bubble.
Stock prices are far higher than actual sales and profits can account for. So one way or another price and value will have to come back together.
While it is not impossible that some breakthrough will lead to a big burst of productivity gains and growth, it is more likely that stock prices will fall.
Regards,
Bill Bonner