The A.I. "Bubble"
If you are in a train and you want to get off you have to start breaking a mile before your destination. A train has a lot of momentum and it can’t be arrested quickly without severe risk to the train, track, surrounding environment and you. But, it also means that if the bridge is out half a mile ahead you have to first, notice the bridge is not there, and then decide when to jump all the while knowing the train and anyone left on it is about to suffer a precipitous drop with a sudden stop.
If you try to get off too soon you’re going to end up bruised and battered and get to watch the train cars and all the other passengers look down on you confused at what you had just done. Really no matter when you jump you are going to do so with apprehension. Because the immediacy of that decision seems more real than the inevitability.
Fail to get off in time and you’ll end up dead. But, you will be free of glassy eyed frowny-faced judgment and ridicule.
This happened in the Titanic. Why jump from the solid firm deck of the ship just because it is sinking?
This is the Buster Keaton version of an equity crash. A train is of course an asset just like a stock. You ride it for a time happily. A train serves a financial purpose since it advances you in some way over just the manual labor of walking.
Most people are just passively riding the train looking out the window at what goes by but never look ahead. In fact, the father back in the cars you are the harder it is look ahead anyways.
The train ride may continue forever, and this is what everyone hopes - since everyone paid for the trip, but occasionally there is a bridge out ahead.
What if there were a dozen trains on tracks at the same time all careening towards bridges that were out but all close enough together you could wave at the other riders. Even if one train engineer saw the bridge and pulled the brake could one train stop safely enough without derail the trains next to it?
That is the current state of the A.I. economy. Even if you are not invested in any A.I. company. Even you are standing on the sideline just watching eventually you’re still at risk of someone pulling a brake and derailing everything.
You have heard the expression “too big to fail” before. Well, the Financial Crisis of 2008 still took out the the economics of Greece and Iceland. I was in Iceland after the fact to play soccer in a tournament there and to take the grand tour of Route 1. I checked into a hotel near the town of Höfn where locals were gathering to listen to a member of parliament give a “it will be fine” speech.
They came prepared.
Prepared with baskets of rotten tomatoes and other other vegetable projectiles. The speaker ran from the conference room looking like a salad.
These weren’t people who had lost money in the market crash. These were just the people wiped out by the derailing trains. The year before the crash economy had been booming. I remember complaining about how expensive a beer was in Reykjavík. That year, the year of the spoiled tomatoes, it was a cheap holiday.
I also got to pet these little ponies who had beards full of icicles.
Right now, the trains are running full speed and everyone is laughing at Michael Burry who just jumped and is still bouncing down a hill while people laugh.
How do you make a train go faster? You can invent a new more powerful engine or you can just get rid of the dead, unnecessary weight - save the braking system.
One choice is expensive and time consuming the other is not. The safety system for a publicly traded company is the cash in the bank account.
In the stock market there are lots of ways of decreasing the “weight” of a company. The easiest way to speed up a stock is just use the cash to buyback the company shares.
And, you can always just order a new safety system from the bond market that will appear in the future. Hopefully, before you need it.
And, you can always tie yourself to other trains. Two engines are better than one, right?
Why stop at two?
What people see is just what is immediately outside of their little window seat.
The bridge is missing. This is a fact. It is in view on the horizon. The people at the front of the train are already jumping off they just are then running into the woods so you don’t notice.
There are only three potential outcomes. All the trains crash. They can’t stop at this point. The only way forward is to fill in the chasm i.e. people start paying to use the tools OpenAi, XAi, Anthropic, etc.
Based on current cost estimates this means every employed person in the US either as a person or by their employer on their behalf starts paying a minimum of $16/month to these companies or based on more realistic projects (this is the depreciation Burry is talking about) ~ $40/month. Or, the 2% willing to pay for these services start paying $800/month ($2000/month.)
This would need to happen at the same time A.I. is causing layoffs of entry level college educated workers.
Is this doable? I don’t think so not unless the Fed steps in and turns back on the money printer.
This narrative ignores the scary side of this. The fact the trains we are on are really just façades of trains. The momentum is just because we are on a hill not because of any actual engine. The fact the engineers studied philosophy and has no idea where the brake lever is and has built in a giant ejector seat they can ride all the way to the edge and then float away.
I would explain the technical in detail. But, I don’t need to.
Read this and this. And listen to this great AI generated voice over. They cover it pretty well.
We have PUTS on the US now.
-AJ