@GaryMarcus I am a thermofluids professor in mechanical engineering. The solution is not physically reproducible. We learn nothing about the nature of turbulence from the result. It’s a mathematical curiosity.
It fulfilled the letter of the clay question, but completely misses the spirit.
Everyone has lost their minds. If you are going to claim that a statistical regression of all of the world’s data is going to destroy humanity, describe the logical progression.
A statistical model has no agency or intelligence. It’s just a relational database of information.
So there are no scenarios where a database rises up and ends humanity without some human(s) on the back end deciding to allow it. These humans would be responsible for crimes and should just be prosecuted like every other criminal.
The problem is not a really big scary database, the problem is human criminals that are not held to account for their crimes.
A few common macro posts that come up over and over on this platform that you can mostly ignore.
1) “The Fed injected [X billion]…”
That’s usually the Fed recycling their maturing securities back into more of the same securities. Or the Fed doing some repo activity, adding a tiny bit of liquidity that comes back out the next day. This year, the Fed is growing their balance sheet very slowly.
2) “The Treasury bought back [X0 billion] of government bonds…”
The Treasury performing buybacks on its own debt is interesting because the deepest, most liquid capital market in the world shouldn’t need the issuer assisting with liquidity. And they could potentially shorten duration over time by issuing t-bills to buy back T-bonds, so that’s worth monitoring. But for the most part, these announcements are nothingburgers. The Treasury issues fresh liquid bonds to buy back aging less-liquid bonds. No appreciable impact on your money or investments.
3) The Treasury has [X trillion] in debt to refinance over the next 12 months, how are they possibly gonna find so many buyers?”
The vast majority of the Treasury’s debt that will be refinanced over a given year will be bought by the same entities holding it now. They’re holding t-bills, those t-bills mature into cash, and they buy the next t-bill. Money market funds, insurance companies, pensions, individuals, corporations, etc. Posting about gross refinance numbers is popular because the number is big and sensationalist, but it doesn’t mean much, which is why you hear about it for years and it’s fine. Net new issuance (about $2 trillion per year) is mostly what matters, and nothing stops that train.
4) “Banks have [X00 billion] in unrealized losses…”
Yeah that was a problem for about one quarter in spring 2023. It’s mostly a non-issue since then but whenever the quarterly number comes out, people love to report it like it’s devastating new information. The number is trending flat-to-down, and it’s a small portion of bank assets and equity capital.
Anyway, good morning. Carry on.
What have I been wrong about, other than the depths of delusion the cult of AI would elicit? In fact, my predictions that it would have no positive impact on productivity, and would actually be inflationary due to high cost and lack of productivity, have been spot on.
From a technical perspective, I have been correct as well. All of the economic outcomes stem from the fact that this is all just elaborate brute force, and 3 years in the only tool they have beat on is more and more brute force.
The last prediction, that it cannot compete with human intelligence economically is slowly coming true as bond yields expand into junk territory and people are getting wise to the round tripping all of the hyperscalers and megacaps are doing to sustain the delusion of real demand.
Keep in mind I started making all of these predictions in 2023 and early 2024, based simply on physical thermodynamic grounds.
The entire time I was pretty clear that bigger models with more data will naturally be able to emulate more things. This of course has happened. What I did not expect was the lengths to which we were willing to incinerate money to try and brute force things that are already done more efficiently.
And I want to be clear that this will all absolutely unwind with lots of pain. And the more money we throw at it, the worse it will be. There is no fairytale ending where really smart people figure it out. Statistical brute force is an economic dead end as a replacement for knowledge work.
Not your keys, not your coins. Also even if they are your keys, they’re not your coins. All your wealth exists in a quantum state of being both your coins and not your coins depending on whether you rolled a dice correctly 99 times and put the right number into your little tamagotchi thing
⚠️ WARNING
Every global crash starts with South Korea.
It sounds ridiculous.
Until you look at the history.
1997:
- The Korean won collapsed
- KOSPI crashed
- Foreign exchange reserves nearly ran out
- Then the Asian Financial Crisis spread across the world
2000:
- South Korean semiconductor stocks topped first
- Months later, Nasdaq dumped 78%
- The Dot Com Bubble popped
2008:
- Korean stocks started breaking down
- Just over a month later, Lehman Brothers collapsed
- The global financial system went with it
2021:
- Bill Hwang used extreme leverage through Archegos
- The fund exploded
- Global banks lost more than $10 billion
- Credit Suisse alone lost $5.5 billion
2022:
- Do Kwon launched LUNA and UST
- More than $40 billion vanished almost overnight
- Then came Three Arrows Capital, Celsius, and FTX
Now the warning is coming from KOSPI again.
Samsung and SK Hynix dominate the index.
More than $30 billion flowed into leveraged products tied to the two companies.
These products rebalance every day.
When prices rise, they buy more.
When prices fall, they are forced to sell.
The deeper the market falls, the more selling the products create.
KOSPI has already fallen 25% in just a month.
The leverage that created the boom is now creating the crash.
Forced selling is feeding more forced selling.
South Korea has warned the world before.
It's warning us again.
Pay attention.
"Howard Lutnick lost 658 Cantor Fitzgerald employees who were killed on 9/11..."
Within days — before many were even confirmed dead — Howard Lutnick cut off their paychecks to their families without warning.
He wept on national television. The families were furious.
Years later Cantor Fitzgerald won a $135 million settlement from American Airlines.
Lutnick told senior executives the bulk of the money would go to the firm's partners — not the families of the 658 dead.
He is now Trump's Secretary of Commerce.
He is also the man who lied to Congress about visiting Jeffrey Epstein's island with eight children.
About 40% of American voters are abject imbeciles, unable to discern reality from fiction— believing the earth is flat, vaccines are dangerous, they’re “eating the dogs,” etc.
If you can’t tell that Trump is by far the most stupid person ever elected, you’re clearly one of them.
Dear $EBAY Shareholders,
Proposal 4 is about your basic rights.
Right now, it takes 20% of all shares to call a special meeting. Proposal 4 lowers that to 10% so shareholders can act when the board will not.
GameStop and Ryan Cohen now hold about 9% of eBay and have offered $125 per share and a plan to increase EPS for shareholders immediately, which is a 40% premium from when they first bought eBay stock in February and a significant premium to today’s price, valuing the company at an all time high since going public. The board rejected this offer while themselves selling shares at much lower prices and, to my knowledge, no current director has EVER purchased eBay stock on the open market with their own money, only sold what they were given.
If Proposal 4 passes and GameStop reaches 10% ownership, they can call a special meeting and you can vote on who sits on the board and runs this company.
For too long eBay has been run by overpaid executives and serial delegators that have put shareholders and customers last. Vote Yes on Proposal 4 at the eBay annual meeting on June 17, 2026. The board should answer to shareholders and not the other way around. Compensation should not be risk free. There should be accountability for bad decisions that hurt the company financially.
Disclaimer: I am a shareholder of both GameStop and eBay and may benefit if Proposal 4 passes and Ryan Cohen gains greater influence over eBay. This message is for informational and advocacy purposes only. It is not investment advice, a recommendation to buy or sell any security, or a solicitation of proxies. Please review eBay’s official proxy materials and make your own voting and investment decisions.