When AI Hype meets Reality

Investment Analysis Clubs Macro Economic Trends and Risks
1

The WSJ is publishing articles almost daily about the immense spending on AI, the lack of customer end-user purchasing of the service and the incredible debts needed to build out the vision for the future. Here is the latest. I think that speculators are finally getting the message that the AI stock bubble may be a true, classic bubble with no near-term profits.

Since the Magnificent 7 stocks are now 40% of the cap weighted S&P500 index many people who simply invest in a low-cost index fund will be heavily impacted if (when) the bubble bursts.

https://www.wsj.com/tech/ai/when-ai-hype-meets-ai-reality-a-reckoning-in-6-charts-bf8043b4?mod=hp_lead_pos3

When AI Hype Meets AI Reality: A Reckoning in 6 Charts

Record capital expenditures and data-center planning run up against the ground truths of physical infrastructure

By Christopher Mims and Nate Rattner, The Wall Street Journal

Here’s a way to evaluate the riskiness of the world’s collective investment in artificial intelligence: Can we even build all the necessary physical infrastructure? And if so, will the resulting AI-powered products generate enough revenue to pay back that investment?…

There is effectively “infinite” money available right now to build out new data centers, says Jim Schneider, a senior research analyst at Goldman Sachs. All this investment has translated into record spending on the stuff that goes into data centers—aka “AI supercomputers”—all those chips, servers, HVAC systems, transformers, gas turbines, power lines and power plants.

There are absolute, physical limits to how quickly all of that can be delivered. As a result, some projects are already being delayed….

image
image1143Ă—800 22.9 KB

Then JPMorgan Fundamental Research calculated how much new revenue these companies will have to generate to justify that $5 trillion investment in AI supercomputers [through 2030]. The result: AI products would have to create an additional $650 billion a year, indefinitely, to give investors a reasonable 10% annual return. That’s more than 150% of Apple’s yearly revenue, and a far cry from OpenAI’s current revenue of about $20 billion a year….

The takeaway: The projections of AI companies and their partners don’t reflect shortages of equipment. At the same time, these projections assume a gargantuan market for AI-powered products and services. Analysts can’t agree whether that market will materialize as quickly as promised…. [end quote]

The AI bubble resembles the railroad bubble of the 1840s (UK) and 1870s (US) and the internet bubble of 2000. Eventually these technologies became widely used and transformative. But the early investors lost their shirts due to over-investment before the paying customers showed up.

Wendy

8 Likes
2

AI is a bubble.

AI is long-term helpful.

Both statements can (and are) true. History shows us this. This time is never different.

4 Likes
3

This result seems to be for the totality of AI investment. History (Pareto) tells us that 20% will win while 80% will lose. That means that this result has no practical use in stock picking.

The way to recoup the investment and make a decent profit is to exploit Increasing Returns which in practice means using the AI in as many units of product sold as possible, like a few billion iPhones.

  • Can the advertising industry (Meta, Google) do it?
  • Can self driving cars (Tesla, Waymo) do it?
  • Can humanoid robots (Tesla, etc) do it?

I believe this is a better way to approach investing in AI.

The Captain

4 Likes
4

Once again it’s about figuring out who the winners will be and picking the right stocks.

AI is a blinking yellow light. AI is clearly not over. But who will make money at it?

2 Likes
5

It’s too helpful. It did my son’s 4th grade math worksheet in 2 seconds……

6

It helped my high conflict, abusive ex son in law to bamboozle a domestic violence evaluator into determining he was the victim of coercive control and financial abuse (as documented just about a year ago). Looks set to be long term helpful too in his extensive and extended, post divorce harassment campaign.

1 Like
7

Will Uncle Sam’s deep pockets be the answer to AI’s prayers?

A perennial characteristic of Silicon Valley startup companies is that they lose a lot of money, at least at first. That’s what happened to Amazon, Uber, YouTube, etc. But to my knowledge, no tech company has ever burned more cash more quickly than OpenAI.

In 2024, it lost about $5 billion; in the first half of 2025, it lost a reported $13.5 billion; and in the last quarter alone, it lost another $12 billion. For artificial intelligence to ever pencil out, some truly enormous revenue streams will be required—$2 trillion by 2030, according to Bain & Company. As the company at the center of the AI boom (along with Nvidia), OpenAI would represent a sizable chunk of that money.

Faced with this dilemma—where do you get a trillion dollars quick?—OpenAI is getting ready to run hat in hand to the taxpayer for subsidies, like every great Ayn Randian self-created entrepreneur, pulling themselves up by their bootstraps. At a recent Wall Street Journal tech conference, OpenAI Chief Financial Officer Sarah Friar suggested that a government loan guarantee might be necessary to fund the enormous investments needed to keep the company at the cutting edge.

1 Like
8

Depends on your investment window. Most here seem to be retired and depend on stock income, so the investment window probably isn’t that long. In other words, if it takes the 20% winners 10 years to recover from the crash and you need the money during that time, you will take a hit even if you picked right.

Case in point, I think we all would agree that Amazon is a dot-com winner. It took Amazon 10 years (2009) to reach its 1999 peak after the dot-com 2000 crash.

Maybe. What I haven’t seem much are estimates of what is the cost of AI. If a CEO wants an AI good enough to replace 50% of employees, how much will that cost in software fees, robot maintenance/repairs, IT, cybersecurity, etc? In my experience, the more complicated the software, the more glitches occur. Without those numbers, how does one calculate competitiveness with human employees? Given the amount of energy and infrastructure needed, I’m guessing the AI annual subscription fee will not be cheap.

Furthermore, if companies are using the same AI, what differentiates the companies? For example, why choose Morgan Stanley advising over JPMorgan if both are using the same AI for their analysis?

Human intelligence comes in a great variety of different flavors, which is why an Amazon, Walmart, Coca Cola or Apple can standout if it gathers the right group of humans. How many different flavors of AI will there be? If there are only going to be a couple of dominant AIs that everyone uses, what happens to market competition?

4 Likes
9

The above assumes that all will crash but 20% will recover. Did all crash in 2000 before some recovered? If so, you are right. I don’t have the data to know.

Amazon is a fantastic story of evolution with an outstanding founder & CEO. Amazon online sales were strong enough for Amazon to survive, but when did Amazon introduce AWS?

Google AI:

Amazon introduced Amazon Web Services (AWS) in 2006, initially with the launch of its Simple Storage Service (S3) in March and followed by Elastic Compute Cloud (EC2) in August. This launch was an expansion of its internal infrastructure services into a commercial cloud offering for businesses and developers.

2003: Amazon began building the internal infrastructure that would become AWS.

An online book store does not qualify as an Increasing Returns business. AWS does. In 2000 Amazon would not have qualified.

Replacing employees does not qualify as an Increasing Returns strategy

One flavor, neural networks. Thousands of applications. Some will create Increasing Returns, others (replacing employees) won’t.

The Captain

1 Like