The A.I. Magic Show Is Ending - What Now?

The A.I. magic show is ending and the audience is learning much of what they witnessed in the “A.I. Boom” is going to look a lot like some other bubbles.

This bubble is likely to be bigger and have far reaching consequences - because when the A.I. collapse comes - which it will as OpenAI cannot deliver the revenue to support its costs and valuation - you, the citizen will be left behind.

Sure, investors will get burned, but the biggest victims may well be the thousands of families who had their land virtually confiscated for hyperscaler data centers.

Here is a snippet from Ed Zitron, an outstanding tech analyst as well as an award winning podcaster:

The question is whether OpenAI can pay the rent, and that’s where things get really messy for those pushing these massive hyperscaler data centers.

OpenAI is a retail phenomenon in software - which is pretty new.

In the past, when cool new technologies hit the market, they were invisible to the common person, the 401(k) investor, the person on the street.

Who ever heard about the revolution in client server architecture or object oriented whatever - so whether they were viable or not was a tech discussion where revenue and adoption would make the call.

ChatGPT is a completely different animal - it’s a retail thing where anyone can ask any question and get a sophisticated answer. So the ChatGPT guys did what all MBAs and mid level venture capital people do - they built a spreadsheet with a line pointing up and to the right - forever.

Revenue would explode - because 100 million people started using ChatGPT pretty early - so it had the same adoption as the iPhone - or some thought.

Nobody was willing to question ChatGPT - because the tech analysts are pretty much in the bag for the major software companies - particularly Oracle who is heavily leveraged on the success of ChatGPT.

So, up and to the right the demand curve shot!

This handy little snippet from Angelo Kourfakas, Senior Investment Analyst for Edward Jones reflects that the 45 degree line is maybe only a couple of degrees up to the right - if at all.

Making matters better the always behind-the-curve Wall Street Journal - who touts story after story about how A.I. will take a large part of the U.S. workforce to the unemployment line chimed in:

We could post lots of articles here but there is something you, the reader, need to discern - in what direction are the stories pointing?

There is no way you can approach the A.I. battlefield right now and come to the conclusion these hyperscaler data centers will ever be supported by the demand - and that is why they are going to be stranded assets - like JC Penney strip malls, in a shorter time that even we predicted.

This is the data center - the hyperscaler data center - in your neighborhood in the next couple of years:

When we make these arguments, as we have for 3 years - and as we did at the Texas State Senate hearing two weeks ago (right after Tanya Tucker spoke - very cool) - the first question we get is:

“So how can you guys be so smart when the richest guys in the world are investing billions of dollars in data centers?”

Well, the answer to that fallacious question is best given by our new favorite analyst Ed Zitron again. Here is the notice all those smart guys - at Oracle - had to put in their financial filings:

And why might it NOT pay off?

Here’s why:

Remember, those hyperscaler data centers are not for your local A.i. guru - they are for ChatGPT and Anthropic - and those guys aren’t bringing in the dough.

What the politicos, the builders, the land developers, the tech analysts sold everyone was that ChatGPT kind of demand was going to change the world. It would be used by everyone, all the time for about any question.

The enthusiasm was - and remains - unbounded - so the scam can get more and more data centers built before the bottom falls out.

What many miss is the A.I. industry is pretty much only two companies - The ChatGPT guys and the Anthropic guys - and if they are not getting the dough - the revenue type, they cannot sustain themselves.

People, the analysts are not forecasting Oracle - one of the largest investors in data centers - is going to the moon. They are predicting when it is going to collapse, its bonds going below junk grade and Larry Ellison may have to sell Hawaii to make a margin call.

Here is a piece from S&P Global - who is in the business of risk determination:

OpenAI makes up much of the A.I. industry - like half.

If OpenAI hits the wall, and the numbers appear now that crash and burn are in the immediate future - Oracle goes down with it.

So, the next time one of the questions we receive is - “well why is all the smart money, the billionaires with all their experts - building data centers?” we can respond because they are the same people who told other billionaires Kodak did not have to get into instant photography or Blockbuster did not need to check out that streaming thing.

One of the refrains we often hear is young people, the digital native, the 30-year-old who grew up with a digital device from almost birth - would support the A.I. explosion.

That is one of those things that sounds good but isn’t really true. Take a look at this Pew Research poll showing younger people do not even like A.I. and may not adopt A.I. for everything - like the forecasters hope:

So that iPhone adoption curve A.I. fanatics claim is going to be the same for A.I. because young people are the future consumers - may not be happening.

We suggest as you investigate the emerging stories about how A.I. is being deployed you delve into why so many analysts and thought leaders are cutting the cards and backing off.

All tech busts come from the single source - lack of revenue to support sky high valuations.

The A.I. bust is coming soon when OpenAI collapses and takes so many supporting players down with it.

That happens when two things precipitate it:

There are gazillions of sunk costs spent on infrastructure to run OpenAI based on wildly optimistic revenue forecasts - and

The rev does not come even close to what is needed to support the story.

It appears those two dynamics are now in place.

Unfortunately for the poor souls who had their land confiscated in Virginia, Georgia, Wisconsin and other states for hyperscaler data centers - the politicians can only say:

Oops!

The Black Swan Files publishes the information you will not read anywhere else - certainly not in the tech press, the Wall Street Journal, The MIT Tech Review - because they are in the pocket of the obsolete tech giants who NEED centralized data centers.

We were among the first - years ago - showing why A.I. economics would not subsidize massive, centralized data centers.

We showed how the U.S. Government failed - and continues to fail - to be able to manage drone swarms - when all comms is lost - and only the Chinese and a couple of small software companies can do it.

Our team currently serves as expert witnesses in 5 states, where we demonstrate huge, data center-consuming applications - running on a computer you can hold in your hand - an Apple mini.

Please subscribe at: TheBlackSwanFiles.Substack.com because we are just getting started delivering industry experts who challenge the false narrative that the future of A.I. is massive data centers.