TRIGGERnometry

TRIGGERnometry

Guest Spotlight

Steve Eisman

Investor.

Triggernometry's avatar
Triggernometry
Nov 19, 2026
∙ Paid
Transcript: Steve Eisman on the 'Paradigm Shift' Happening in the Stock  Market - Bloomberg

The financial crash of ‘08 was as devastating as any single event this past century. Its impact is difficult to overstate; nearly 20 years on, its malaise continues to inform our politics, society, and sense of well-being. Yet, the slim few who saw it coming were able to avoid the worst of it. Some of them even got rich.

Steve Eisman is one such case.

Rising to fame when he shorted collateralized debt obligations, Steve became one of the most talked-about financial advisors in the world.

In 2015, he was portrayed (though by the pseudonym ‘Mark Baum’) by Steve Carell in the Oscar-winning comedy The Big Short, an adaptation of the 2010 book of the same name.

Last year, Steve launched a weekly podcast - The Real Eisman Playbook - where he extols behind-the-scenes financial advice, strategies, and insight.


Why we invited him on.

How do we know when we’re looking at a bubble? For most of us, we don’t. Not until it bursts. Then it becomes radically apparent, and none of us can believe we didn’t see it sooner.

Housing was one. NFTs were another. The dot-com craze and tulips. Bubbles are nothing new, but there are always new ones coming.

Today, we might be living in one of our own - AI.

Those who love and loathe it agree on one thing: it’s going to change everything. It may destroy the world’ it may deliver us to utopia. It may be terrible, but it’s certainly inevitable. Whichever flavour of prospecting you opt for, it’s a ripe investment opportunity. Best case scenario, you’re in on the ground floor of a technology that will make the entire industrial revolution look like a sand castle. Worst case scenario, everyone’s dead anyway.

But what if they’re both wrong?

That’s what we want to find out.


What we talked about.

”Nobody serious believes that LLMs or agentic AI are ‘thinking.’ It’s nonsense. They’re highly sophisticated predictors - LLMs predict, using all their data, what the next word in an answer is going to be, and agentic AI predict how to perform a task. They’re not ‘thinking’ - they’re predicting."

It sounds like a distinction without a difference. Isn’t that how human brains work? A series of receptors and synapses fire along long-established pathways to initiate effective courses of action. Where is predicting unique from thinking?

”AI has no ability to think critically. It’s just a software program predicting the next word. It predicts the next word very well, but it’s not the same as thinking … It’s a very subtle difference. Critical thinking requires making judgements. Using the information you have to make moral, rational judgements. It also can’t discover anything. ‘Predicting the next word’ is going to have very powerful implications for global finance, so I don’t want to understate that, but I also don’t want to overstate it.”

And those ‘very powerful implications’ are a source of much debate at the moment. AI and its development are the single most significant driving force in finance today; what the technology does and doesn’t achieve will determine the allocation of trillions upon trillions of dollars. Whatever comes of it, a wealthy sector of society has decided to put it all on black.

”The tech world has concluded that this new technology is going to change the world, and they cannot allow themselves to be left out. If they get left out, they will be destroyed. They must participate it. It turns out, that participating is extremely expensive … The computation costs are huge, so there’s a lot of money to be made. The chip makers, the hyperscalers [who build the data centres]… They’re getting involved.”

Long before these investments are set to actually pay off, the bills being run up are utterly astounding. The computation that AI needs is a paradigm shift from what we already have, even in our immensely technological world, and the hardware this software necessitates comes with a hefty price tag.

”We’ve always had data centres. But the computational power required by AI is so huge that the kind of data centres we have now just won’t cut it. The ones we’re building now are five times the size of any we’ve ever had, and every single one of them costs at least $5 billion, in some cases more. Just this year alone, those hyperscalers - Amazon, Google, Oracle, etc. - will spend $850 billion on building them. Next year, they’ll spend over a trillion.”

Trillions of dollars spent in the pursuit of… what exactly? A brave new world, or the end of ours? Even board members from within the very companies developing the technology warn of 50% chance of total annihilation of the human species. What does Steve make of these bets?

User's avatar

Continue reading this post for free, courtesy of Triggernometry.

Or purchase a paid subscription.
© 2026 Triggernometry · Privacy ∙ Terms ∙ Collection notice
Start your SubstackGet the app
Substack is the home for writing, podcasting, and media founders.