AI Is Distorting Reality
More from Jackson Hole... Could AI 'distort' the markets even more?... A solution Kevin Warsh probably likes... The latest on the labor market isn't great... What really sets long-term bond yields...
Central bankers have been warned...
Artificial-intelligence ("AI") agents might soon know policymakers' decisions before they know themselves...
That's what Princeton University economist Markus Brunnermeier says. On Saturday, Brunnermeier presented an academic paper on AI and monetary policy at the annual central bankers' confab in Jackson Hole, Wyoming.
AI dominated a lot of the conversation last weekend, with Federal Reserve Chair Kevin Warsh talking up the technology's growth and productivity potential, as AI investments are already significantly boosting GDP numbers.
Brunnermeier went in a different direction... bringing a big elephant into the room. And it's making some headlines. As Brunnermeier argues in his paper...
AI agents can learn how humans think and respond, while humans may be unable to understand or reliably anticipate how those agents will act.
He basically posits that machines, like the kind that support instantaneous trading on Wall Street, could become good enough at modeling central bank and market behavior that they effectively know what the Fed is going to do before the bank has finished deciding it.
And that fact would distort market prices well before the central bank makes a decision or public comment. Brunnermeier calls this "asymmetric understanding." You could call it robots replacing humans.
The solution? Less is more...
Now, I (Corey McLaughlin) don't think AI models can consistently and accurately predict surprise "emergency" maneuvers from the central bank.
The models are always right... until they aren't. Then some overleveraged firm is exposed, contagion emerges, and the game changes.
And the market has already been "distorted" for years because of the anticipation of central bank policy – not to mention actual policy itself. So that's nothing new.
But I get Brunnermeier's main idea about considering the further rise of machines in investing. And more importantly, Brunnermeier's suggestion for dealing with this scenario might be welcomed by the Fed. That is, less is more.
Brunnermeier says central banks should think about being less predictable to outsmart AI models and temper them from front-running Fed decisions that impact stable prices and the labor market. Brunnermeier writes in the paper...
With asymmetric understanding, transparency has to be rethought as predictability arms the opponent in the financial dominance game.
I wasn't there, but I can imagine Warsh nodding at that point. "Less, or no, guidance" has been his MO since becoming Fed chair. And the expectation is that there might be fewer Fed press conferences and less signaling in the future.
As I've written before, the approach is likely to increase market volatility tied to Fed decisions (or indecision) and allow the central bank to be less accountable. But now, Warsh can say a Princeton economist thinks it's a good idea, too.
We've seen a similar concept playing out across industries...
With the rise of AI, companies are seeing the value in protecting their data "sovereignty" and preventing AI models from getting their hands on businesses' unique advantages.
As for real jobs...
It's a busy week for labor-market data. Yesterday, the Bureau of Labor Statistics ("BLS") released its Job Openings and Labor Turnover Survey ("JOLTS") report for July.
There were 7.27 million job openings across the economy in July – up slightly from June's downwardly revised reading of 7.18 million. Hiring during the month fell to 5.1 million, hitting the third-lowest level since the COVID-19 pandemic.
There hasn't been a lot of change in the JOLTS report in recent months. But it's important to look at the broader trends...
Job openings have been between 6.5 million and 7.5 million for more than two years now. And the hiring component of the JOLTS survey has been in a steady downtrend since the post-COVID hiring boom.
But one number really stands out: Hiring, or lack thereof, in "professional and business services" – a category that covers staffing agencies, consulting firms, and computer-systems design. This category saw hiring decline by 188,000 jobs in July, which could be an important signal. Temporary staffing and consultant work are some of the first to go when various companies start bracing for a slowdown.
This morning's data from payroll processor ADP showed more of the same...
The private sector added 38,000 jobs in August. That was below the estimate of 48,000 job gains and July's reading of 46,000 added jobs. August marked the third straight month of slowing job gains and had the fewest jobs added since January.
Only five of the 10 sectors ADP tracks added jobs in August, with education (back to school) and health services leading. And the manufacturing sector lost jobs again (17,000), with the manufacturing workforce now at its lowest level since December 2021.
Businesses are still hiring, but the pace of those hires is slowing.
While inflation is the Fed's clear focus right now, the labor market is softening...
On Friday, the BLS will release its nonfarm payroll data for August.
That's the most important release of the week. In July's data, the labor market lost jobs for the first time since February. The market will be hoping for a bounce back in August. But if we see continued weakness in the labor market, coupled with high(er) inflation, it may change the Fed's priorities again – or at least complicate them.
A weak labor market means the Fed may not raise rates this month like the market has been expecting. So inflation could continue to run hot moving ahead.
That's our human take (take that, AI agents!), and it hasn't been priced into the market yet. Federal-funds futures traders only marginally decreased their bets today on a hike at the Fed's September 15-16 meeting, from 67% to 60%.
Here's another real take...
Yesterday, Stansberry Research senior analyst Alan Gula released a must-read issue of The Total Portfolio.
In it, Alan analyzes Treasury Secretary Scott Bessent's recent announcement on the "Treasury Twist," designed to "silence the bond market" and influence long-term yields, which harkens back to Bessent's trading days on Wall Street. As Alan notes...
He's effectively trying to lower long-term interest rates by having the Treasury become a buyer of long-term bonds. (More demand for bonds drives prices higher, which, in turn, lowers yields.)
In 2020 and 2021, the Treasury issued bonds with remarkably low coupons – including a 2% bond due in 2051, sold at par (quoted at $100). Today, that bond trades for around $54.
Selling at $100 and buying back at $54 seems like a good trade. But keep in mind that the government would be swapping out one debt obligation for another at a higher interest rate. So interest expense will rise.
Alan also shares his thoughts on famed investor Stanley Druckenmiller's public (and AI-aided) takedown of Bessent's plans – before getting into what you really need to pay attention to in the bond market.
As Alan writes...
What actually drives long-term yields?
Above all else, short-term rates and nominal economic growth.
That means the Fed – which largely influences short-term rates through its policy – drives the economy. Given all of this, Alan reaches a clear conclusion about what the bond market is really signaling – and the answer might surprise you.
In any case, Alan says, "The Total Portfolio is prepared for whatever might come at us. In fact, many of our positions are flying following Bessent's Treasury Twist and its not-so-surprising secondary effects."
We can't give away all the details in these pages, but Portfolio Solutions subscribers and Stansberry Alliance members can find all the details here.
New 52-week highs (as of 9/1/26): Alpha Architect 1-3 Month Box Fund (BOXX), Chord Energy (CHRD), iMGP DBi Managed Futures Strategy Fund (DBMF), Equinor (EQNR), Cambria Emerging Shareholder Yield Fund (EYLD), Helmerich & Payne (HP), Kayne Anderson Energy Infrastructure Fund (KYN), Marathon Petroleum (MPC), Motorola Solutions (MSI), Plains All American Pipeline (PAA), USCF SummerHaven Dynamic Commodity Strategy No K-1 Fund (SDCI), Valero Energy (VLO), and State Street Energy Select Sector SPDR Fund (XLE).
In today's mail, feedback on yesterday's Digest and mail – where we noted high(er) inflation once again... plus a question about our Stansberry Conference coming up later this month... Do you have a comment or question? As always, e-mail us at feedback@stansberryresearch.com.
"Wages are rising. Inflation is declining. Why are you guys always so negative on the U.S. economy?" – Subscriber Glenn S.
Corey McLaughlin comment: Talk about painting with a broad brush...
"Will there be a livestream option for the 2026 conference?" – Stansberry Alliance member Jim W.
McLaughlin comment: The short answer is yes. We'll be sending out more information about our Livestream Pass over the next few weeks. In the meantime, you can call 1-800-201-4147 to get set up with access now.
All the best,
Corey McLaughlin with Nick Koziol
Baltimore, Maryland
September 2, 2026