Non-Technological Constraints to AI

Market manias have patterns. The most powerful ones are genuine technological revolutions pushed far beyond rational limits by crowd psychology.

By mid-1999 it was already clear to veteran investors and students of economic history that the dotcom bubble had reached parabolic insanity.

The speculative momentum was still unstoppable – and would run a lot further – but grown-ups knew by then that few of the high-flying start-ups were ever going to generate a viable revenue stream. The authentic success stories would have to fight each other in a cannibalistic struggle for survival.

We are nearing the same point today with AI, although this time for a different and overwhelming reason. The $20tn (£15tn) valuation of hyperscalers, chipmakers and the larger AI complex, has wildly outpaced the electrical infrastructure needed to run data centres and sustain the technology on anything like the projected scale.

The physical constraint is rock hard. “Our grid in the United States hasn’t had any meaningful upgrade since the 1970s,” said Bobby Majumder from the industrial law firm FBT Gibbons.

The threat to AI stock mania is not so much lack of energy – though that is serious – but rather the global bottleneck of transformers, substations, switchgear, transmission lines and all the unsexy stuff we rarely think about, leaving aside the acute shortage of skilled workers in the US able to install and run such kit.

A single big campus in the data centre hub of Hays County, Texas – an area where I once played a lot of golf (misspent youth) and know well – can use 10 million gallons of water a day for evaporative cooling and power generation, draining the Edwards Aquifer that also supplies the Austin-San Antonio corridor.

“Nobody is talking about cooling; nobody is talking about water,” said Majumder, speaking at the recent Marshall & Stevens forum on energy infrastructure. “The farmers are not going to be happy at all about you pumping down their aquifer for cooling.”

There are other obvious catalysts that could puncture the bubble. Stubborn US inflation – input prices are rising at the fastest pace in four years – may force the Federal Reserve to stop its “stealth-QE” via bill purchases. The bond markets may hold Kevin Warsh’s feet to the fire as he takes over the institution.

Inflation may stop Scott Bessent, the poacher turned gamekeeper now running the US treasury like a hedge fund, from using the $8tn money market to help soak up massive fiscal deficits at the peak of the economic cycle.

Cheaper “commoditised” AI from the likes of DeepSeek in China may start to undercut American rivals, threatening the implicit pricing model behind today’s equity valuations. If it is true that DeepSeek v4 can achieve 80pc-90pc of the performance of Anthropic’s Claude at 10pc of the cost, you start to see the problem.

Liaquat Ahamed, author of the wonderful Lords of Finance covering the Great Depression and now releasing his new book 1873, likens the AI boom to the American railway mania after the Civil War. Routes were duplicated in the rush for dominance.

Costly lines passed through sparsely inhabited regions where there would never be enough human traffic in time to justify the scale of debt issuance. [...]

Hyperscalers can try to leapfrog the grid bottleneck by building their own power plants, but that will not solve the problem either, at least not in time to alleviate the burden of fast-mounting and opaque AI debt.

It took 17 years to plan, license and build the recent Vogtle nuclear plant in Georgia. Costs ballooned from $12bn to $30bn. Small modular reactors may be cheaper per gigawatt – don’t hold your breath – but none yet exist in the West, and there will be no serious supply chain until circa 2040.

Shale gas frackers can drill until they drop, but that makes no difference if there are no gas turbines available on the world market. The waiting list for heavy-duty models used in combined-cycle plants has stretched to seven years, although hyperscalers with the deepest pockets are jumping the queue for a fat fee with 2030 delivery dates. [...]

The AI revolution is real. The language models are fabulous. The technology will make economic life almost unrecognisable by mid-century.

But the internet revolution was also real in 1999 before the Nasdaq index dropped 77pc, flushed out the commercial nonsense and overshot in the other direction.

Don’t track Nvidia chip orders if you want to know where the AI market is heading. Track the metaphorical picks and shovels that make it all possible. 

by Ambrose Evans-Pritchard, The Telegraph | Read more:
Image: Richard Newstead
[ed. See also: How bad is AI for the environment? (Yale Climate Connections).]

Cadillac Desert

CADILLAC DESERT: The American West and Its Disappearing Water. By Marc Reisner. Illustrated. 582 pp. New York: Viking.

It's unlikely that most taxpayers will read ''Cadillac Desert: The American West and Its Disappearing Water,'' but they should. It's a revealing, absorbing, often amusing and alarming report on where billions of their dollars have gone - and where a lot more are going.

The money has gone into Federal water projects in the Western states - some of the projects awesome, some scandalous but all with an uncertain future. More than a century ago John Wesley Powell, the nation's pioneer hydrographer and an explorer of the Grand Canyon, concluded that so much of the West was virtually desert that if all the flowing water in the region were applied to it, the water would spread too thin to make much difference.

But that didn't daunt several generations of pioneers, who believed the selective harnessing of available water could yield miracles. And it did. It virtually created modern California, making it the nation's most populous state and one of the world's prime agricultural areas. On a smaller scale, similar marvels were wrought in other states - Arizona, Utah, Colorado, the Dakotas, Montana and even Nevada.

It all came about less through engineering skill than through political prestidigitation. There's a thing known in Federal circles as the Iron Triangle. One side - depending on the week - is either the Interior Department's Bureau of Reclamation or the Army Corps of Engineers, rival bureaucracies dependent for their existence on the building of dams and related water facilities. The second side of the triangle consists of members of Congress, shamelessly wooing votes via pork-barrel projects. On the third side are beneficiaries of water projects - farmers, contractors, merchants, local politicians and a host of secondary opportunists. Link these together, and you have a greed machine, fueled by taxpayers, that for generations has been unbeatable. President Carter tried to challenge it with his ''hit list'' of questionable water projects and came out of Congress's threshing machine too battered to swing a second term.

The taxpayers' problem is that the chronicle of this hocus-pocus normally emerges in inconclusive bits and pieces, in reports based on sanctimonious handouts from the Bureau of Reclamation and the Corps of Engineers that are heavy on how they are saving the world, light on what it's costing - and often opaque about the justification for the projects.

Marc Reisner, a former staff writer for the respected newsletter of the Natural Resources Defense Council, has put the story together in trenchant form. He details the Machiavellian competition between the bureau and the engineers, recounts how huge sums have been spent to benefit small numbers of influential people and suggests painful days of reckoning lie ahead.

Parts of his account are oft-told stories, such as Los Angeles's snaffling of water from farmers 300 miles away. But much of his material is fresh and powerful, taken from such previously unplumbed sources as the bureau's ''blue envelope'' (secret correspondence) files and a marvelous, hair-down interview with Floyd Dominy, its free-swinging former commissioner. The 1976 collapse of the Teton Dam in Idaho - an instance of a structure that never should have been built - is detailed for the first time, with all its implications of carelessness and incompetence. Mr. Reisner also makes clear that much Western irrigation has been based on reckless ''mining'' of water in the great Ogallala Aquifer, which extends into seven states, from Texas to South Dakota. The severe depletion of this eons-old unrenewable resource, he says, has been matched in other areas by a reckless indifference to the accumulation of salts in soils. This has killed farmland and caused drainage crises like the current mess at California's Kesterson Reservoir, where pollution has poisoned the wildlife.

''None of this,'' Mr. Reisner writes, ''is to say that we shouldn't have gone out and tried to civilize the arid West by building water projects and dams. It is merely to suggest that we overreached ourselves.'' He maintains: ''What federal water development has amounted to, in the end, is a uniquely productive, creative vandalism. Agricultural paradises were formed out of seas of sand and humps of rock. Sprawling cities sprouted out of nowhere. . . . Its worst critics have to acknowledge its positive side. . . . The cost of all this, however, was a vandalization of both our natural heritage and our economic future, and the reckoning has not even begun. . . . Who is going to pay to rescue the salt-poisoned land? To dredge trillions of tons of silt out of the expiring reservoirs? . . . Somewhere down the line our descendants are going to inherit a bill for all this vaunted success, and . . . it will be a miracle if they can pay it.''

by Gladwin Hill, NY Times |  Read more:

Image: via

[ed. A classic, and the bill's about to come due.]

Clearing the Market

Cushing, Oklahoma is the pricing point for West Texas Intermediate (WTI) crude and the physical hub through which US oil supply flows to refineries across the Midwest and Gulf Coast. As of 25 June, inventories have fallen to 19 million barrels, below the operational minimum (~20mb) that the industry considers the threshold for physical stress. The US Strategic Petroleum Reserve (SPR) has fallen to 331.2 million barrels, the lowest since 1983. According to the IEA, global inventories are at their lowest seasonal point in recorded history.

The market has priced almost none of this. 

- Two Spikes Coming?

***

• The ceasefire holds, but the underlying deal has stalled on the points that determine whether reopening is sustainable.

• Iran is entrenching control over the strait through mechanisms – mines, fees – that outlast any ceasefire.

• Iran’s institutions can’t agree among themselves, so even a signed deal still may not compel the IRGC, hence the physical reopening the market is pricing isn’t coming on the MOU’s own timetable.

• Meanwhile the price is being held down by three cushions – released barrels that had been trapped in the Gulf, SPR drawdowns, and Chinese reserves and reduced imports – that are all finite, so the mispricing identified in Two Spikes Coming hasn’t resolved.

• New evidence this week – inbound tanker numbers, floating storage, operator testimony – confirms the physical picture rather than the price picture.

The argument in Two Spikes Coming rests on a race, between stockpile depletion and production restoration, with Cushing already near its operational floor and reserves elsewhere running out within one to two months. The past week has not changed that race, but it continues to indicate the reserve draws are still outpacing the return of flow.

The MOU signed on 17 June was supposed to settle the reopening. Instead it has settled into a pattern of brief traffic windows followed by a strike, a US response, and a return to the negotiating table to manage the aftermath. The Ever Lovely was hit on 25 June inside the safe corridor the International Maritime Organisation (IMO) and Oman had set up along the Omani coast days earlier. The US struck Iranian missile, drone, and radar sites the next day. A projectile then hit the tanker Kiku and Iran fired at US positions in Bahrain and Kuwait. The US next expanded its target list to surveillance, communications, and minelaying infrastructure. Both sides then turned up in Doha this week and kept communicating, which suggests the ceasefire itself is intact, but does not indicate the deal underpinning it is progressing.

The two sides are not talking directly to each other in Doha. American and Iranian delegations are meeting Qatari and Pakistani mediators separately, a step back from the direct sessions held in Switzerland two weeks earlier. Iran’s stated priority is Clause 11, the release of frozen assets, and President Pezeshkian has stated that $6 billion of the $12 billion held in Qatar will be returned, though it is not yet clear on what terms or even whether the funds have moved. Iran will not discuss its nuclear programme until those funds move, but if that money moves without a matching concession on enrichment it will reduce the leverage the US has left for that discussion. Trump has claimed a deal was close at least 38 times between late March and early June, according to a CNN count. Doha is yet one more round in that pattern.

The strait’s governance is where the deal has stalled most. Iran’s foreign minister Araghchi has said the removal of “obstacles” in the strait, and its reopening, rests with Iran alone. The IMO’s Secretary-General has said Iran laid an estimated eighty mines across the main shipping channel. The timetable for clearing them is set by Tehran regardless of what Doha produces. The Joint Maritime Information Centre raised the strait’s security threat level to “substantial” this week, citing mine risk and clearance uncertainty. Oman has separately delivered a service-fee proposal to Washington and its allies. An Iranian official has called the fees mandatory, but a regional diplomat has called them voluntary. Either way, Iran can prioritise the shippers who comply and delay the ones who do not, so the dispute over wording matters less than the authority it establishes.

Inside Iran, more than sixty of the Assembly of Experts’ roughly 88 members signed a statement on 28 June warning negotiators against crossing Khamenei’s red lines, control of the strait among them. The Assembly’s own secretariat publicly distanced itself from the statement within hours, which means even the body meant to speak for Iran’s clerical establishment cannot agree on how hard a line to take. Pezeshkian spent the same week in Qom telling senior clerics the opposite, that the MOU was an economic win worth defending. And while the president was making that case, the IRGC struck a vessel inside a corridor the foreign ministry had just endorsed. Three arms of the same state, pulling three different directions, in the same seven days. No single part of the Iranian state can bind the others to one position, hence why incidents the negotiators did not authorise keep recurring.

WTI is trading around $70, close to its level before the war began, and Morgan Stanley has cut its Brent forecast on the basis that Hormuz is reopening faster than expected, projecting a 2027 surplus of 4.8 million barrels a day. “Strip away the narrative,” the bank’s analysts wrote, “and read only the prices. They describe a market that has weakened across the board.” Morgan Stanley may be right about a near-term glut – outbound cargo has genuinely surged since the MOU – but the mistake is extrapolating that burst into durable recovery. The analysts are getting the direction backwards because a weak price does not necessarily prove a weak market. Rather, in this case it means a market distorted by reserve draws and supply disruption, and both of those aspects are temporary props under the market, not foundational features. That makes Morgan Stanley’s case much harder to sustain.

Roughly 170 million barrels of crude that had been trapped in the Gulf cleared the market once the MOU allowed it out. The SPR is drawing at a pace that leaves perhaps three to six weeks of room. Chinese crude imports have fallen by something like 5 mb/d since March, while China’s visible commercial stocks have barely moved, which means the shortfall is being met from reserves that do not appear in any published series. Cushing itself fell to 18.96 million barrels in the week to 19 June, the lowest since October 2014 and near the roughly 20 million barrels traders treat as an operational floor. A partial reopening of the strait does not fix that on its own.

by Nick Wade, State of Play |  Read more:

Image: EIA, HFI Research

[ed. See also: Trump Paused War to Manipulate Oil Prices (video/YT). And, this: New Report Reveals True Extent of Devastation of US Fifth Fleet HQ in Bahrain with a special status update on F-35 readiness. Priceless. (Another good(?) read here).  I don't think this war is going the way they thought it would.]

The Licensing Revolution: Is Resistance Futile? Part 1: Loss of Ownership Comes to the Car

“Neoliberalism as an economic system enshrines the extraction of rent over industrial production.”
—Yours truly, here

Two of the most revolutionary inventions man ever made were created in the 20th century, one at its start and the other close to the end. Both offered the same innovation: a quantum advance in individual freedom and power.

I’m talking, of course, about the automobile, personal transportation, and the PC, your own personal computer.

Cars and Computers

If you own a car, you own your own transportation; you don’t rent it or borrow it. You can argue the merits of “owning” personal transportation — there are climate, pollution, and crowding arguments against — but there’s no question about the freedom it gives to people. You want to leave now? Just jump in the car and go.

If you own a PC, same thing. Before the PC, some calculations and modeling were just too painful and time-consuming to do, and many were simply impossible. Think of the most complicated spreadsheet you’ve ever created — could you have done that by hand? Or better, if you could have done it by hand, would you have?

Before the PC and its business equivalent, the UNIX-based Sun Workstation, access to computing power were through IBM-style mainframes and minicomputers, like those made by DEC. None of these could be considered “personal”; they were too costly, and though they could accommodate multiple users at terminals, the computing itself was centralized and corporate-owned.

Keep this in mind: Before the PC, computing was centralized and corporate-owned. After the PC, computing power was inside the box you worked at, and priced for individual sale. Now thanks to Windows 11, that’s all been reversed.

Cars and computers, each a revolution in personal power and control. Now both will be taken away. Your car will no longer be yours, nor will your PC.

Soon You Won’t Own Your Car

The above statement is true in too many ways. The car you’ve already bought will be licensed to you, a license that can be revoked.

Your New Car Is a Spy

Cars have become computers over the last few years. And that means cars have become spy machines. Here’s one review, by the Mozilla Foundation, of the automobile industry from the standpoint of privacy, written in 2023. Its bottom line is the headline:

It’s Official: Cars Are the Worst Product Category We Have Ever Reviewed for Privacy

All 25 car brands we researched earned our *Privacy Not Included warning label -- making cars the official worst category of products for privacy that we have ever reviewed.

The link for individual brand reviews is here. Their sins are many; these are the important ones:

1. They collect too much personal data (all of them)
2. Most (84%) share or sell your data
3. Most (92%) give drivers little to no control over their personal data
4. We couldn’t confirm whether any of them meet our Minimum Security Standards

Recipients of the sale of your data could include your insurance company, which can purchase everything recorded about your driving habits.

And you can’t shut this stuff off, because it’s not hardware, but software, and the car needs its software to run. Here’s Tesla’s warning about its software, again from 2023 (emphasis mine):

However, “if you no longer wish for us to collect vehicle data or any other data from your Tesla vehicle, please contact us to deactivate connectivity. Please note, certain advanced features such as over-the-air updates, remote services, and interactivity with mobile applications and in-car features such as location search, Internet radio, voice commands, and web browser functionality rely on such connectivity. If you choose to opt out of vehicle data collection (with the exception of in-car Data Sharing preferences), we will not be able to know or notify you of issues applicable to your vehicle in real time. This may result in your vehicle suffering from reduced functionality, serious damage, or inoperability.”

It’s gotten worse since then; Tesla’s just getting started.

The Biden Bill–Mandated ‘Kill Switch’

Watch the Breaking Points video at the top; it details, from reputable reporters, the next dystopian “feature” of cars manufactured in 2027 and later — a “kill switch” that turns your car off if it thinks you shouldn’t be driving.

The detail is here. Basically, under Joe Biden, Section 24220 of the Infrastructure Investment and Jobs Act “requires all new passenger vehicles to eventually include factory-installed technology that detects driver impairment and prevents or limits vehicle operation.”

The implementation falls under the NHTSA, which is writing the rule. Barring congressional prevention or modification, the kill switch is expected appear in all newly manufactured cars (but not used ones) starting in late 2026 or early 2027.

Privacy and Control

In modern America, two things are certainly true. 1) Once privacy is taken away, it never comes back; and 2) when a power is gained by corporations and government, they pervert it as fast as they can.

The prime example is this war — because Congress long ago surrendered its war-making power, the Executive has steadily moved in, to the point that today there’s not even a pretense of getting congressional permission. Trump wants a war wherever, that’s what he does. Or consider the definition of “terrorist” — today it’s “whomever the feds wishes to hurt, and to whatever extent.”

So what’s the maximum harm that can be done by the “new automobile”? Your driving is monitored by AI; the data is fine-grained and stored; anyone who wants it can buy it for whatever goal, including to raise your insurance, or deny you coverage.

Further, anyone with control of the software — the manufacturer, the FBI (initially under subpoena, but later, who knows?), cops, Homeland Security, or any branch of the law, whatever that means — can turn off your car when it wants, or (why not?) gain full control, lock you in, and drive you wherever it wishes. Remember, eventually every new power is perverted.

It starts, as always, with calls to Save the Children (MADD is mad for this law).

The next expansion is to further the War Against Crime. (“Remember the OJ Simpson highway chase? What if they could just turn off the car? You want to catch OJ, right? Do you hate the cops?”)

Then it transforms into … what? Whatever the security state wants, because “keeping you safe.”

The Licensing Revolution

You won’t own your car for another reason as well. You may have noticed a trend: what you used to be able to buy, you now merely rent.

• Apple doesn’t sell music, it licenses use.

• You no longer own your software. TurboTax, for example, sells a “personal, limited, nonexclusive, nontransferable, revocable license to use the applicable Software only for the period of use provided in the ordering and activation terms”.

• Same with Amazon’s ebooks and audiobooks.

• Same with Microsoft Windows. (More on that later.)

Non-transferable and revokable licenses. Renting your life.

by Thomas Neuberger, God's Spies |  Read more:

Images: uncredited; and Branimir Kvartuc/ZUMAPRESS.com/Corbis

[ed. In fact, Sony just made news the other day about remotely deleting all previously purchased content in digital libraries, and a couple days later ditching physical disks in favor of licensing. Amazon has already made this transition with Amazon Prime videos. See also: The Licensing Revolution: Windows Edition. Part 2: The computer you bought isn't yours. A tale about power.]

***

"Words have meaning. Proper word selection is integral to strong communication, whether it’s about relaying one’s feelings to another or explaining the terms of a deal, agreement, or transaction.

Language can be confusing, but typically when something is available to “buy,” ownership of that good or access to that service is offered in exchange for money. That’s not really the case, though, when it comes to digital content.

Often, streaming services like Amazon Prime Video offer customers the options to “rent” digital content for a few days or to “buy” it. Some might think that picking “buy” means that they can view the content indefinitely. But these purchases are really just long-term licenses to watch the content for as long as the streaming service has the right to distribute it—which could be for years, months, or days after the transaction." via:

The Billion Dollar Crypto Man

President Donald Trump took in nearly $1.2 billion dollars from his crypto businesses last year, a federal filing released Tuesday shows, locking in profits while his investors were socked with losses.

Mere startups when he took the oath of office, the new ventures have now eclipsed in revenue much of his vast property portfolio that took him decades to accumulate. Fueling their rise were billionaire investors and Trump’s own move to quash a federal crackdown on the industry.

Trump got more than $500 million from his World Liberty Financial business selling new crypto products, including “governance tokens,” according to the required annual disclosure report with the Office of Government Ethics. It also showed another crypto business, CIC Digital LLC, took in more than $600 million from sales of souvenir-type “meme” coins stamped with his face.

Both the tokens and the coins have plunged in value since the sales.

Trump also took in millions last year from selling Trump-branded bibles, sneakers and other small items in another unprecedented move for the presidency. The sale of Trump-branded watches alone brought in $4.7 million.

The 927-page disclosure form paints a stark, if incomplete picture of the massive growth of the president’s wealth since taking office last January through a web of business interests — many that have benefited from the policy moves of Trump’s own government. Trump has insisted that his sons direct his finances but the arrangement rejects the conflict of interest protections that his recent predecessors in office had instituted.

Forbes estimates Trump’s net worth at $6 billion, up from $2.3 billion in 2024.

The Trump business is growing abroad

The rise of crypto relative to Trump’s property is especially noteworthy because he first rode to office boasting of his property wins. It’s also remarkable because that mainstay business also boomed last year. Trump took in tens of millions in fees from a flurry of new hotel, resort and condo deals overseas that amounts to the biggest property expansion ever in the century since the family business was founded.

Many of those countries were negotiating with the U.S. over tariffs, military aid, and other important matters.

A property in the United Arab Emirates took in $10.4 million. One in Saudi Arabia being built by a real estate developer close to the ruling family sent the president’s company $9 million. And one in Bucharest, Romania, and another in Qatar sent him $5 million each.

One of his prominent domestic properties, Mar-a-Lago in Florida, notched big growth last year, too.

Trump took in in $77 million from the property, a 50% jump from the year earlier when he was just another citizen, as heads of state and business people flocked to it in his new term.

The disclosure report doesn’t give profit figures, just revenue, so it’s impossible to know how much he is earning.

Trump is now the billion dollar crypto man

After taking office last year, Trump reversed the Biden administration’s tough stance on the crypto industry and pushed policies friendly to the industry.

But regulators still had some concerns. Before Trump’s World Liberty began selling “governance tokens,” they issued warnings about this new kind of crypto asset, saying that unlike stocks, the tokens offer no ownership stake in the issuing company, just voting power on certain corporate polices, and are difficult to value.

Buyers pounced anyway, including a Chinese billionaire who spent $75 million on the tokens and $200 million on the souvenir coins. In February last year, a federal lawsuit charging him with duping investors was paused before being settled last month for a $10 million fine. [...]

Meanwhile, investors have seen the value of their meme coin holdings drop significantly. The price spiked to more than $74 in the days after its launch in January 2025, but now sells for just $1.68. Also, the value of the World Liberty tokens has fallen 80% since they first started trading in September.

by Bernard Condon, Seattle Times/AP |  Read more:

Image: Alex Brandon
[ed. This actually plays like a feel-good story. The sheep MAGA cultists and influence buyers get fleeced - as predicted, as they deserve (Under the Trump crypto playbook, the family always wins. Investors don’t). Is this a great country or what? In other corruption news, see also: Trump is using a $500M no-bid contract to build his White House ballroom (Washington Post):]

***

White House officials last year secretly awarded a no-bid contract worth up to $500 million for the construction of the East Wing ballroom in an unusual arrangement that sidestepped typical contracting procedures designed to control costs, according to a copy of the agreement obtained by The Washington Post. [...]

The estimated East Wing construction cost has tripled since July, when the project was first announced, with half expected to come from taxpayers, The Post previously reported.

Trump has repeatedly claimed that the ballroom would be paid for by private donors and once said that Clark executives offered to build it for free.

“They said: ‘Sir, we’ll do it for nothing. This is the greatest honor,” Trump told The New York Times in January.

Clark’s internal cost projections show the McLean, Virginia-based company, the largest general contractor in the D.C. metro area, stands to make tens of millions of dollars from the work...

The records reviewed by The Post do not break out Clark’s estimated profit margin for the entire project, but a March document shows the company projected it would receive a total of $65 million in combined profit, overhead and daily rates for on-site staff and other costs.

[ed. But, but... Hilary's emails!]

What's Elon Worth?

[ed. Let's see, by my watch $103,000 every minute. He could own every major league sports team 2.5 times over; buy all the gold in Ft. Knox (with billions to spare)... (more).]

via:

Meta Culpa

Early last year, Meta's chief technology officer, Andrew Bosworth, had a clear message for his staff. "You should quit if you feel that way," he told one employee who said workers were being treated poorly. "You should consider working elsewhere," he told another person who questioned controversial changes at the business. He was reinforcing the company Meta had spent the last few years trying to become: a lean, fast, high-pressure organization that no longer had the patience for internal debate. "You can leave," Bosworth said, "or disagree and commit."

But this month, in a memo and a meeting with employees, Bosworth sounded like a different person. Morale is "probably one of the worst it's ever been," he said, adding that the business had done "an atrocious job" with its recent restructuring. "We've undermined the trust you have that your specific expertise and contribution will be valued."

Since 2022, Meta has remade itself around a ruthless management playbook that helped define a new era in Silicon Valley. Through relentless layoffs and many other unpopular decisions, executives charged ahead, emboldened by record profits and apparently immune to the building discontent. Bosworth's comments last week were different — an acknowledgement that Meta's leadership may finally be confronting the costs of its actions.

Meta's workforce is at a breaking point. Employees in the UK are trying to form a labor union, decrying executives' "cruel and shortsighted behaviors." More than 1,600 workers have signed a petition demanding that Meta stop tracking employees' keystrokes to improve its AI models. As Wired reported this month, things have gotten so bad that one frustrated employee hijacked a livestreamed meeting with a profanity-laced outburst directed at an executive. Another compared working in a new AI-training unit to the gulag. Others are so dejected they're actually praying to get laid off so they can leave with at least some severance.

Against this backdrop, Bosworth was one of several executives in recent weeks scrambling to do damage control. Chief Product Officer Chris Cox acknowledged the "insanity of this company" that created a "difficult" and "brutal" environment. CEO Mark Zuckerberg admitted "we've made mistakes."

"It's a classic example of chickens coming home to roost," says Sandra Sucher, a professor of management practice at Harvard Business School. "They have almost systematically destroyed trust. They are trying to figure out how to dig themselves out of the hole that they dug."

The digging started with a mass layoff of 11,000 people in late 2022, which Zuckerberg was at least apologetic about. The company then slashed another 10,000 jobs the next spring in what Zuckerberg hailed as a "year of efficiency," and then another 3,600 in 2025 that he said was to get rid of "low performers," effectively torpedoing some workers' job searches (many of them, it turned out, had received good performance reviews). In March this year, news leaked that the company was about to ax even more jobs, but it didn't confirm the cuts for weeks and didn't notify those affected until May, sending everyone into a nauseating, two-month purgatory. In April, amid the limbo, Meta announced it would start tracking employees' keystrokes, stoking fears that the company wanted to automate their work. And in May, as it laid off 8,000 employees, it reassigned another 7,000, many of them to menial jobs that involve training AI. Meta declined to comment on this story. [...]

For employees caught in the hailstorm, it must have felt validating for an executive to empathize with their situation. But surely he and the rest of Meta's leadership knew all these things would make employees unhappy, and yet they did them anyway. So why the sudden mea culpa?

Perhaps all the anger, dissatisfaction, and open rebellion was harming productivity. Or the particularly public nature of Meta's dysfunction, with the crescendo of news reports, had become a liability for its reputation with investors. Or maybe executives finally realized what had become patently obvious to everyone else — that whatever Meta was doing just wasn't working. The whole point of adopting this hard-charging management style was to get employees to innovate faster and catch up to competitors like OpenAI, Anthropic, and Google in the all-consuming battle over AI. Instead, Meta has been falling farther and farther behind.

by Aki Ito, Business Insider |  Read more:

Image: Wally Skalij/Getty; Getty Images; Tyler Le/BI
[ed. Why anyone would want Facebook/Meta on their business resume is beyond me. The money might be good, but working for a company like that would just be burning life years. See also: The Internet Has Become Too American to Trust (The Walrus).]

Bitcoin Winter

Why bitcoin is trading at a 2-year low

Bitcoin is in the dumps.

The apex cryptocurrency is down more than 30% in 2026, slumping to its lowest level since 2024 this week to trade around $59,200. That marks about a 53% drop from the token's all-time high above $126,000 last October. Ethereum, the second-biggest crypto, is doing even worse, down 48% this year.

So, how did the Trump-era bullishness for crypto that took bitcoin to record highs give way to a brutal crypto winter that shows little sign of thawing?

Scanning around, there's little good news to be gleaned from the headlines. Interest from institutional and retail investors alike looks tapped out. That's evidenced by record outflows from bitcoin ETFs, which Deutsche Bank says have hit $6 billion in six weeks, the longest losing streak since the funds were launched in early 2024.

"This matters because ETF demand has become central to Bitcoin's price formation: the same vehicles that supported the 2024-25 rally now amplify the decline mechanically when flows reverse," the bank wrote this week.

And then there's Strategy. The business data firm founded by Michael Saylor, the "never sell" bitcoin evangelist, has sold some bitcoin.

"On 1 June, Strategy (formerly MicroStrategy) disclosed it had sold 32 bitcoin, its first sale since December 2022. The amount was negligible (0.004% of holdings) and the immediate price reaction was modest (~3%), but the signal was not," Deutsche wrote.

In the days following the sale, bitcoin dropped nearly 20%.

Selling by Strategy—which is the largest corporate holder of bitcoin—has been a looming question mark all year as the token's price has slumped below the Strategy's average cost. The difference in the firm's net asset value of its bitcoin holdings versus its market cap has driven speculation that it could sell more tokens, an event that would weigh further on sentiment.

"Bitcoin currently trades below Strategy's average cost of $75,699, and the market has begun to price the possibility of forced selling by leveraged corporate holders. We expect this question to persist," Deutsche analyst said.

Capital is also flowing away from bitcoin and into another speculative area of the market: AI. Analysts attribute much of the waning enthusiasm for crypto among retail traders directly to their relentless appetite for artificial intelligence. The rapid outflows from bitcoin ETFs in the last month have been mirrored in blistering pace of investing in many of the top AI and chip ETFs.

A final wrinkle is the surprisingly hawkish new Fed boss, Kevin Warsh. The policy meeting this month officially dashed all hope of a rate cut, with Warsh's first meeting further boosting odds of a rate hike, a bearish development for risk assets like bitcoin.

by Max Adams, Business Insider |  Read more:

Image: Yahoo Finance; Jennifer Sor/BI

What It Means to Be a Democratic Socialist

“To me, what socialism means is to guarantee a basic level of dignity. It’s asserting the value of saying that the America we want and the America that we are proud of is one in which all children can access a dignified education. It’s one in which no person is too poor to have the medicines they need to live.”

Alexandria Ocasio-Cortez, Vogue, 2018

Democratic socialists’ decisive congressional victories on Tuesday night in New York’s primary elections solidified the far-left movement as an ascendant power center in blue states.

Now, as the progressive coalition prepares to expand its footprint in Washington, many Americans are turning their attention to the movement for the first time — and wondering, perhaps, what it actually stands for.

The definition often depends on whom you talk to. But the movement’s standard-bearers are united by their belief that direct government action — not the free market — is a better tool to solve problems for everyday Americans, such as the rising cost of health care and housing.

“Economic stress is something I lived with as a kid, and I feel it in my guts,” Senator Bernie Sanders, independent of Vermont and an architect of the movement’s modern resurgence, said in an interview with The New York Times. “That’s what makes me a democratic socialist.”

In the United States, democratic socialists’ policies tend to support working within the capitalist system rather than abolishing it outright. Critics typically decry the likely high costs to taxpayers of some of these policies.

Ashik Siddique, a co-chairman of the organization, said the group surpassed 100,000 members earlier this year. About 1,000 more joined after the sweep of victories in New York on Tuesday night, he said.

Here is a closer look at the pillars of democratic socialism.

End Military Aid to Israel

The defining feature of primary races in New York on Tuesday was a litmus test on American support for Israel. Democratic socialists won that ideological battle handily, since staunch opposition to continued military aid is a key part of their campaigns.

The Democratic Socialists of America, a political organization in which members pay dues and are organized around a wide-reaching policy platform, says it “stands for the full freedoms and self determination of the Palestinian people, including the end of Israel’s colonization and occupation of all Arab lands, equality, and the right of all refugees to return to their homes and properties.”

Mr. Sanders said every time he has talked about Gaza at rallies across the country, he has received a standing ovation.

Expand the Social Safety Net

Democratic socialists want the government to lower the cost of living for Americans. Under their platform, child care, pre-K and public higher education amount to a collective good and should be completely free and funded by the government. They also support universal rent control, and want every worker to receive paid family leave.

In New York City, it was the political machine of Mayor Zohran Mamdani, a democratic socialist, that helped carry three progressive House candidates to victory on Tuesday.

Mr. Mamdani plans to open a free preschool center on the Upper East Side. Although directed at working families, the move has ignited a fierce debate over whether a city facing a major budget deficit should use taxpayer money to fund a free service in affluent neighborhoods.

Guarantee Free Health Care

The D.S.A. wants to create a single, government-run national program providing essential health care for everyone.

Right now, individuals and employers pay insurance premiums. People pay cash co-payments for drugs. And state governments pay a share of Medicaid costs. The system is expensive, but it allows individuals some choice in their care.
In a democratic socialist system, like one long trumpeted by Mr. Sanders, nearly all of that would be replaced by federal spending.

Many democratic socialists want to see private insurance entirely eliminated. Others are open to giving people the option to keep their private insurance plans.

Tax the Rich

There is no consensus about how much such a system would cost the federal government, nor exactly how it would be funded.

Proponents of democratic socialism say that higher income taxes on wealthy Americans and decreases in military spending would cover the costs.

by Emily Davies, NY Times |  Read more:

Image: Graham Dickie for The New York Times
[ed. An over-simplified and somewhat dismissive description of DSA policies, but at least this political philosophy is finally getting some attention. See also: ‘American Democratic Socialism’ Has a Proud, Diverse, and Inspiring History; and, The Left is Rising (Currrent Affairs); and Why the DSA and socialists are on the rise now in US cities (Vox); also Wikipedia's definition: Democratic Socialism.]

***

I'm going to hold off on any 'irrational exuberance' for now, but if there's one slogan I'd suggest any DSA campaign use, it's: "You own government. Make it work for you." That, after all, is basically the central theme of democratic socialism. DS isn't some monolithic political philosophy, with entrenched political policies. It's not Russia or China. It's an adaptable model, flexible enough to respond to shifting problems and priorities within the dicates of the US Constitution. It doesn't seek to wipe out corporations or any other businesses large or small, but it does want to make sure that there's a level playing field for everyone so that opportunity exists on all levels. The economic benefits produced from this capitalist system not only flow to shareholders, but also back into government programs and public improvements that everybody can benefit from and enjoy (like infrastructure). The worst thing (which opponents always glom onto) would be to focus too much on cultural issues or granular details (eg. appropriate levels of policing and incarceration; gender issues, etc.) and letting the big picture get lost in the weeds. Let those things play out in courts, not political platforms. It's time for change. New generations are crying out for it, and one benefit of the Trump years is that there's now a new understanding of what's possible in terms of shifting boundaries (and what tactics can be used). We need a new direction and DSA is the best option I've seen.]

[ed. Update: Again, establishment Democrats continue to shoot themselves in the foot, and provide more ammunition to Republicans by allowing themselves to be defined by what they're afraid of rather than what they stand for... See: Centrist Democrats Rebuke Party’s Left Wing: ‘We Are Capitalist, Not Socialist’(NYT):]

***

“The bottom line is that you have to give the D.S.A. and you have to give MAGA credit, because they’re organized,” Mr. Suozzi said, referring to the Democratic Socialists of America, the country’s largest socialist organization. “And the people that don’t agree with their philosophies wring their hands at cocktail parties, but they’re not organized. So we have to get organized.” [ed. 'Cocktail party' democrats, a winning message.]

Mr. Suozzi said democratic socialists were tapping into “real economic anxiety” and were “right in their diagnosis of the problem.” But he argued that Democrats should pursue policies grounded not in socialism but in a pro-union form of capitalism. [ed. with unions looking soon to be roadkill on the way to AI.]

A spokeswoman for Mr. Mamdani, Dora Pekec, pushed back on the letter, saying in a statement that the “only thing extreme is defending a status quo where working families can’t afford to live.”

And a representative for the Democratic Socialists of America, Priscilla Yeverino, said in a statement that the group was gaining popularity because it was pursuing policies that Americans support, and that “Red Scare tactics are no longer working.”

“Ending wars, passing Medicare for All, forgiving student loan debt, abolishing ICE and taxing the rich — those are all popular policies” said the statement. [...]

Matt Bennett, a co-founder of Third Way, a centrist Democratic think tank, said the outcomes in New York were “dangerous” for Democrats nationally.

“What we’ve seen Republicans do very successfully before is weaponize the craziest ideas of the activist left,” he said. “And now the ammunition they’ve got is much, much more powerful.”

***

[ed. Update 2: Fortunately, Republicans are even more disorganized and demoralized than democrats, and their "ammunition" mostly blanks. See: Behind the Curtain: The cost of blind loyalty (Axios).]

Are Americans Too Old?

The country you live in is changing. Month by month, year by year, an insurgent group has been taking over. Its members are moving into your neighborhood, casting votes, and pushing your interests aside. These people claim to care about the community, but they’re mostly loyal to one another—and their numbers are growing. If their ascendance has been ignored, that’s mainly because of political correctness: it’s considered rude to talk about them as a group. If you do so, you must adopt a respectful, even reverential tone, observing how hard life is for them, even though they have all the power.

“They” are the old—at least, according to “Gerontocracy in America,” a new book by Samuel Moyn, a professor at Yale Law School. Moyn argues that the oldest Americans, because of their retrograde politics and ever-increasing presence, are profoundly reshaping our collective life. Historically, “elderly Americans have counted among the most oppressed,” he writes, and many still suffer abuse, or struggle in penury. But the bigger picture is that more Americans are living longer, staying healthier, and getting much wealthier as they age. As a result, Moyn says, the country’s fate and character are being determined not by forward-looking people in their youth or their prime but by backward-looking ones in the final third of their lives.

The French have a phrase for stating the obvious: “enfoncer une porte ouverte,” or “to break down an open door.” We all know that there are lots of boomers, and that Joe Biden and Donald Trump are the oldest Presidents in history. Even so, Moyn writes, the extent of America’s transformation has, like aging itself, snuck up on us. His title is a play on Alexis de Tocqueville’s “Democracy in America”: it implies that gerontocracy—rule by the old—is now the country’s essential condition. “Had she won the presidency in 2024, Kamala Harris would have taken office at sixty,” Moyn points out; only in a gerontocratic America could she have presented herself as a youthful alternative.

To really appreciate the “gobsmacking” degree to which the country has aged, Moyn suggests, you have to look at the statistics. In 1980, the median age in America was thirty. (In other words, half of Americans were younger than thirty, and half older.) Today, the median age is nearly forty. There used to be an “age pyramid,” Moyn explains, with a broad base of younger people narrowing to a small elderly population at the top. Now we have an age rectangle—more people are reaching their seventies and eighties—and it could soon become a top-heavy trapezoid, since young people are having fewer children. In 1920, less than five per cent of Americans were older than sixty-five; by 2060, according to the A.A.R.P., the number will be one in four.

The age of the median voter is now fifty-two. In primaries, it is sixty-five—meaning that the oldest voters ordain the choices for the rest of us. “The most common age of donors in recent elections can run as high as seventy,” Moyn reports; since politicians often do what donors want, even younger elected officials are likely to vote older than their age. That’s not to say that there are lots of younger politicians: the median age in Congress is more than sixty. There are four hundred and thirty-five members of the House of Representatives; only one was born in the nineteen-nineties, and only sixty-four in the eighties. Democrats in Congress trend a little older than Republicans, and “at least half of the Democrats in the House over seventy-five are running again in 2026,” Moyn writes, despite the fact that, between 2022 and 2025, eight congressional Democrats died in office.

All of this has made younger voters more cynical and disengaged. And with good reason: there is ample evidence that older people favor policies that emphasize security for themselves over investment in the young. Broadly speaking, laws now make it much easier for older people to buy property and make investments while avoiding taxes. Meanwhile, being healthier, they have kept working into their seventies, occupying positions that might otherwise be filled by those younger than them. The result has been a widening economic rift between the old and the young, with the net worth of older households rising and the wealth of younger households falling. “The age group most likely to own a home in America, at a rate of over 80 percent, is seventy to seventy-­four,” Moyn writes. The second most likely group is people seventy-five and older.

There are nearly sixty million Americans over the age of sixty-five. Can we really generalize about their attitudes and opinions? “As the individual life dwindles, playing for time in the face of impending catastrophe is a psychologically appealing stratagem of avoidance and denial,” Moyn suggests. At the very least, it seems reasonable to say that our opinions grow less au courant as we age. Surveys find that, among people aged eighteen to twenty-nine, the most important foreign-policy issue is climate change; among “old people,” Moyn writes, “the biggest issue is terrorism.” We face all sorts of big civilizational challenges—and yet, if Moyn’s analysis is right, the people who are most directly invested in building the future are being dominated by those who indulge the status quo. “Gerontocracies are prone to let long-term problems fester and worsen,” Moyn warns. But the power of older Americans is hardly despotic; it’s democratic, deriving from the principle of one person, one vote. What, if anything, should be done about it? [...]

Is gerontocracy the right diagnosis for what ails us? In an essay titled “Old People Aren’t the Problem,” Nathan J. Robinson, the editor of Current Affairs, argues that Moyn is making a category mistake. Not all older people are wealthy and powerful; in fact, in 2019, seventy per cent of the wealth owned by those over sixty-five belonged to just ten per cent of American seniors. “Wealth is not actually concentrated among old or young people,” Robinson writes. “It’s concentrated among rich people.” He points out that, in modern America, the politician who has done the most to advance progressive ideas is Bernie Sanders, who is now eighty-four years old (and, to all appearances, totally with it). Would the world be a better place if Sanders were mandatorily retired? “The class struggle overlaps a bit with age, but the policies we should adopt have to be aimed around redistributing wealth and power, period,” Robinson concludes—otherwise we’ll just be “exploited by a younger ruling class.” [...]

The fault lines between young and old are real. I’m in my mid-forties, with two small children, and I live in one of only a few school districts on Long Island where the school budget failed to pass; most of the people I know reasonably assume that it was older voters, wary of even modest tax increases, who voted it down, happy to risk the drastic cuts to programs like tutoring, music, and sports that will occur if a new budget isn’t passed. (On Facebook, there are arguments between parents who want services for their kids and older residents who say those services didn’t exist “back in my day.”) There are vacant lots and empty buildings in town where new housing could be built, but residents, defensive of their property values, keep nixing new development. The status quo rules. And yet it’s not just older people who cling to the past. A mood of retrospection seems to have settled everywhere. In conversation, almost no one will express hope for the future. Maybe one sign that we’re living under gerontocracy is that so many people yearn for the old version of America, in which dynamism abounded and everyone was young.

by Joshua Rothman, New Yorker | Read more:
Image: Josie Norton

[ed. I'm old, and old people drive me nuts. But, the slow, steady transfer of accumulated wealth over the next couple of decades will have a big impact on these issues. Will lucky recipients act any differently?]

Escaping the Ogallala Trap

There is a closing window to stop driverless cars from creating omnigridlock.

Self-driving cars are not a hypothetical future but a familiar part of the urban background in San Francisco. I have driven in them several times and the novelty of seeing a steering wheel turn itself has pretty much worn off. During 2026, Waymo service will expand to Dallas, Houston, San Antonio, Orlando, and Miami, joining Atlanta, Austin, Los Angeles, San Francisco, and Phoenix.

Right now, self driving is a premium experience, more expensive than a human driver, in part because Waymo uses new cars, and in part because there are still relatively few Waymos on the road, spreading operational overheads thickly on a small fleet. Over time, Waymo and its competitors will become cheaper than human-driven taxis.

You make driving fun

Self-driving cars need not look like traditional cars inside. Normal cars are heavy and bulky, in large part due to safety requirements. Despite sharing the road with human drivers, Waymos already have 80 percent fewer accidents. When self-driving cars become 90 percent of the cars on the road, they will be able to platoon and join up into little trains, saving the space usually spent on gaps between vehicles and doubling road capacities.

They can be more comfortable as well. The Volkswagen GEN.TRAVEL has seats that fold out into flat beds, with passenger restraints for safe sleeping while moving and lighting designed to generate natural circadian rhythms. The Volvo 360c offers a first-class private cabin with a classic Volvo touch: a special safety blanket that acts like a seatbelt, usually loose and comfortable but tightening instantly on impact. In theory it can be an entertainment space or a mobile office too. Simpler, working versions of this idea, like the Amazon Zoox, are already driving around Las Vegas and San Francisco.

With imagination, you can see how a wide range of functions could be performed in a car: working, sleeping, eating, and even socializing, effectively bringing back the bar cars once enjoyed by New York commuters to Connecticut. I already buy cans of beer for long train rides with my friends. Train lines created entirely new seaside resorts like Atlantic City in the US, and Heringsdorf, Ahlbeck, and Bansin in Germany. Just imagine the trips people would make with the ability to effectively travel business class in their cars, driving overnight.

Our gridlocked future

Autonomous vehicles are the centrifugal water pump of the roads. Just like the Ogalalla Aquifer, most roads are currently free at the point of use. And just like the Ogalalla Aquifer, they will be overused if we do not charge for the privilege of drawing on them. Anyone who needs to get where they’re going quickly will be stuck in traffic with all the people enjoying a beer, working from a mobile office, or having a nap. There will be total gridlock.

Though taxes on fuel and registering cars are universal across the developed world, imposing charges at the point of use has been trickier. It took New York City 60 years to impose congestion pricing, and it was almost revoked several times along the way. London’s congestion charge has survived, but attempts to extend it out of the very inner core have not. Dutch voters destroyed per-mile charges, the Kilometerheffing, in 2010. Hong Kongers rejected such a scheme in the 1980s, despite an effective trial.

These attempts failed for a range of reasons. But a major one is that they aimed to change the rules of the game for everyone at the same time, creating a lot of people who lost out under the policy while giving them nothing in exchange.

by Ben Southwood, Works in Progress | Read more:
Image: Getty

Students Are Using a ‘Backdoor’ to Attend Their Dream Schools

Justin Helman didn’t get his dream acceptance from the University of Florida. But that isn’t stopping him from pursuing the classic college experience there.

The recent high-school graduate from Park Ridge, N.J., is set to move into a private apartment right by campus. He is enrolling in a UF online program for the first few semesters and paying an extra fee package to access services like the campus gym and student-section football-game tickets. He plans to study at the library, join clubs and might rush a fraternity.

“I’m going to get almost the entire same experience, and the only thing I’m really missing is going into class and dorming,” he said. “To me, it was just almost a no-brainer.”

More students like Helman are discovering there is another way into their dream schools.

Students who don’t get into major public flagships the traditional way are still participating in the social life of these campuses. The small-but-mighty group is moving to college towns, enrolling in online programs or nearby community colleges, living in private housing, joining Greek life, and attending game-day tailgates. The approach is sanctioned by the universities, which are expanding alternative-enrollment programs. [...]

Helman’s UF offer was to the school’s Pathway to Campus Enrollment program, which requires students to start online before transitioning to full in-person status. The program has exploded from about 250 students in 2015 to nearly 3,000 in fall 2024, according to the school’s website.

Helman will share a Gainesville, Fla., apartment with three other PaCE students who are moving from out-of-state, and said he has spoken to many others planning to relocate. He chose the program over traditional acceptances, some with scholarships and honors, including at the University of South Carolina, Seton Hall and University of Tennessee.

“This was his dream school,” said his mother, Maria Debowska-Helman. She added that his tuition would be cheaper than a traditional UF student’s. The optional fee package will cost around $550 for a semester, depending on the number of course credits. [...]

It is also controversial. Some students view these alternative pathways as “a cheat code,” Kraemer said. Some consultants agree, at times pointing to limited major-transfer options and instead pushing students to traditional paths.

by Roshan Fernandez, Wall Street Journal | Read more:
Image: Maria Debowska-Helman

What Women See in Men and Vice Versa: Estimates Based on Sex Ratios and Marriage Patterns

Abstract 

Much of what looks like changing marriage preferences over the twentieth century is actually demographics. Exploiting plausibly exogenous variation in sex ratios across U.S. birth cohorts (1870, 1930, 1950), we jointly identify preferences, match quality dynamics, and the costs of marriage and divorce. Demographics alone explain two-thirds of cross-cohort differences. Women’s premium for older husbands collapsed across cohorts; men’s preferences barely changed. Love that survives its early years becomes permanent, but the odds of surviving fell from 97% to 44%. Divorce costs fell six-fold and depend on life stage. A horse race across behavioral channels shows that the match quality process—not mate-age preferences—is the primary dimension of generational change. Declining divorce costs and fragile match quality are substitutes: either alone fits the data, but together they reveal two independent dimensions of social change. The model validates out of sample on the 1910 and 1970 cohorts.

Introduction 

Much of what looks like changing marriage preferences over the twentieth century is actually demographics. Variation in sex ratios and mortality across U.S. birth cohorts—driven by immigration and differential longevity gains—accounts for two-thirds of the cross-cohort differences in marriage and divorce behavior, with no change in behavioral parameters at all. This paper provides the first joint identification of marriage preferences, match quality dynamics, and the costs of marriage and divorce in a unified equilibrium framework, exploiting the large demographic variation across the 1870, 1930, and 1950 birth cohorts.

The variation we exploit is plausibly exogenous to marriage preferences. The sex ratio at marriageable ages fell from 1.056 men per woman (1870 cohort) to 0.942 (1950 cohort)—a swing from male surplus to female surplus driven by the closing of the frontier, declining male immigration, and faster female mortality improvements. These forces operate through the population’s age and sex structure, not through tastes over partners. They change who is scarce and who must compete in the marriage market, so that equilibrium marriage and divorce patterns shift even if no one’s preferences change. We estimate a dynamic general equilibrium model of marriage and divorce, matching 84 moments of marriage and divorce behavior across the three cohorts. Beyond the aggregate role of demographics, the estimates reveal sharp findings about what people value and how relationships work:

• Women’s preferences changed; men’s did not. Women in the 1870 cohort placed a premium on older over younger husbands large enough to delay marriage by several years relative to a world with symmetric preferences (Figure 3). By 1950 this premium had collapsed to near zero. Men’s preferences over partner age are essentially constant across all three cohorts. The marriage age gap is driven not by men preferring younger (more fecund) women, as Siow (1998) suggests, but by women’s preference for older, more established men—a preference that erodes as women gain economic independence. 

• Love that survives becomes permanent—but surviving got harder. A good match, once achieved, is permanent: the implied duration exceeds the remaining lifetime. But in the 1870 cohort a new marriage had a 97% probability of reaching the good state; by 1950, this had fallen to 44%. The 1950 cohort uniquely allowed recovery from bad matches with 16% chances, generating dynamics that resemble cohabitation. 

• Divorce costs depend on life stage. The middle-age group (“young” in our model) faces the 2 Throughout we consider only opposite-sex couples, reflecting the historical period studied. For historical mortality rates, see Haines (1998) and Arias (2012); for historical patterns of gender-biased immigration to the U.S., see Donato and Gabaccia (2015). 2 highest effective divorce cost—roughly six times the utility value of a standard-deviation match shock—substantially above both adolescents and the old. This generates the age-declining divorce rate profile observed in every cohort. The base cost declined six-fold across cohorts, with a structural break between 1910 and 1930 coinciding with the liberalization of divorce laws and the entry of women into the labor force. 

• Divorce costs and match quality are substitutes. Cohort-specific match quality process alone— without any cohort variation in divorce costs—achieves the same fit as the Baseline specification. Both channels govern marital dissolution, one through the price of exit, the other through the probability of wanting to exit. Combining both yields a further 14% improvement, revealing two independent dimensions of social change: the liberalization of exit and the increasing uncertainty of relationships. 

How can these mechanisms be separately identified? The key is that different moments respond to different parameters, and the three cohorts provide 84 moments under very different demographic conditions. Marriage rates by age for male and female reveal how each sex values partners of different maturities: when the sex ratio shifts from male surplus to female surplus, the scarce sex becomes pickier and marriage patterns change in ways that depend on the preference parameters. Divorce rates and their age profile reveal the cost of exit: the pervasive pattern that divorce declines with age identifies the age-dependent component of divorce costs, because without it the model would predict flat or rising divorce with age. The fraction never married by age 50 disciplines marriage frictions: a high never-married fraction signals that substantial frictions prevent matches from forming. The persistence of marriages—how quickly divorce rates fall with duration—reveals match quality dynamics: if good matches are permanent but medium matches are fragile, divorce concentrates in the early years. The cross-cohort variation in these moments overdetermines the parameter vector. [...]

This paper contributes to the literature on marriage and matching in three ways. First, it provides a framework to separate demographic forces from behavioral responses in equilibrium matching markets. Second, it identifies the dynamics of match quality and the role of divorce costs using variation that is orthogonal to preferences. Third, it shows that much of the long-run change in marriage and divorce patterns can be understood as the consequence of demographic shifts rather than changes in tastes.

by Jose-Victor Rios-Rull; Shannon Seitz; Satoshi Tanaka, PIER/University of Pennsylvania |  Read more (pdf):

Meta’s New AI Unit Is a Total Mess

Someone interrupted a livestreamed, employee-only presentation at Meta earlier this week with an expletive-filled outburst about “being the company’s bitch,” according to a recording heard by WIRED. The individual then asked the people leading the call to write to a specific Meta AI executive and "tell him that he's a piece of shit."

One of the presenters covered their face with their hands, according to a witness. (The speaker could not be reached for comment, and the meeting’s two leaders moved on with their technical talk after asking everyone to mute, though employees commented on the stream about the “spicy” start.)

The incident, which took place on a call open to thousands of employees, reflects growing frustration inside the company’s Applied AI team, which was formed in March to support the work of AI researchers at Meta Superintelligence Labs. Three current employees tell WIRED there is widespread dissatisfaction with how Meta assembled the unit of about 6,500 engineers and product managers and the drudgework they allege they have been assigned to improve AI models. Each spoke on the condition of anonymity because they were not authorized to speak to the media.

“It's literally the gulag,” one of the employees claims. “You have zero purpose in life all of a sudden, you barely interact with anyone, you just have these tasks every week."

Another employee describes some of the tasks—generating puzzles to test how reliably AI models from Meta and other companies can solve them—as easy compared to the software development work they had been doing previously. But the new projects feel menial, and “almost all” employees seem unhappy, they say. “Most people find the work soul-crushing,” the third employee says.

Meta declined to comment for this story.

Applied AI isn’t the only unit where tensions are boiling over and contributing to what workers describe as record-low morale. The company’s AI-focused restructuring, which included 10 percent of the company, or 8,000 employees, being let go last month has generated extra work and stress throughout several divisions, including data center engineering and Instagram, several current and former employees tell WIRED.

Across the company, more than 1,600 employees have signed a petition demanding that Meta stop a recently launched initiative to monitor US employees’ clicks and keystrokes to generate AI training data. (The company has scaled back the program slightly, allowing employees to pause data collection for up to 30 minutes and request specific exemptions).

During a meeting this week open to all employees at Instagram, Meta chief product officer Chris Cox addressed the “difficult” and “brutal” environment created by the “insanity of this company” in the past few months, according to a recording heard by WIRED. Cox applauded Instagram employees for launching features and serving around 2 billion users amid what he compared to “running a marathon in the middle of a hailstorm and then, like, your teammate gets replaced and then we’re recording you.”

“It’s like what the fuck,” he said, drawing laughs, before repeating himself. “It is like what the fuck.” [...]

In an internal memo on Friday seen by WIRED, Meta CEO Mark Zuckerberg acknowledged that recent organizational changes had caused distress across Meta. “Given the complexity of these changes, we’ve made mistakes and will almost certainly make more,” he wrote. “As we navigate this period, I’m also focused on providing as much stability going forward as possible.” [...]

“Talented People”

Zuckerberg’s memo also addressed the allegedly dismal situation in Applied AI directly, referring to the unit by its acronym. He suggested the team was a waypoint, not a destination. “Work like AAI is critical to advancing our models and it lets very talented people contribute to those efforts while we create other roles they can contribute to around Meta over the coming months as well,” he wrote.

Engineers selected for the unit have no choice but to join or leave the company, an unusual requirement for highly valued technical employees in Silicon Valley. That’s led some members of Applied AI to describe themselves as “draftees.”

The organization has grown in batches since early April. “It’s crazy to watch people experience the shock of it as each wave comes in,” an early member of Applied AI says.

Some employees are being asked to finish two tasks per week. These involve generating complex software coding problems to help AI scientists better train and evaluate the performance of the latest frontier models. Some of the work is meant to help develop AI agents that generate software or other outputs.

One worker describes the assignment as “mechanical and not creative,” and certainly “not using their full skill set and knowledge.” They feel they were hired to develop social media apps for billions of people, but now find themselves assembling data for hundreds of AI scientists to feed to computer chips.

Meta released pioneering open-weight AI models three years ago, but has had mixed results with subsequent releases. Applied AI is among several expensive initiatives Zuckerberg has spun up in hopes that the company can better compete in the growing market for AI services.

Zuckerberg noted in his memo that, unlike some other AI labs, “automating work” was not Meta’s primary focus. “The products we’ll build will range from much more personalized Instagram and Facebook experiences and glasses that help you throughout the day to better tools for small businesses to thrive and create jobs, and personal superintelligence agents that understand your goals and work 24/7 on your behalf to help in the ways you want,” he wrote.

by Paresh Dave; Zoë Schiffer, Wired |  Read more:

Image: Kyle Grillot/Getty Images
[ed. Dead company walking. Seems pretty clear (to me, anyway) that they don't have a clue what the company will look like in the future, just that they need to be in the AI space somehow - this after the dismal (and expensive) failure of the company's 'Metaverse' makeover.]