Showing posts with label Economics. Show all posts
Showing posts with label Economics. Show all posts

Wednesday, August 19, 2026

Big Tech’s AI Spending Is $3 Trillion Higher Than It Seems

Each quarter, big tech companies disclose their massive capital expenditures on artificial-intelligence infrastructure, from data centers to chips.

But those figures don’t come close to expressing the full extent of future spending to which Google parent Alphabet, Meta Platforms, Oracle and many others have committed. That is because a huge swath of their coming financial obligations aren’t reflected on their balance sheets.

Nine top tech companies had some $3 trillion of off-balance-sheet commitments mostly related to AI, according to a Wall Street Journal analysis of footnotes in their most recent securities filings. Those obligations are growing faster than traditional “capex,” which totaled about $600 billion over the past year they reported, and were about triple what the companies owe under their outstanding leases and long-term borrowings.


America’s blue-chip tech companies are placing these huge bets based on assumptions about what the demand for AI computing—and availability of AI hardware—will be in several years. Their hope is that they will easily meet all their obligations with future revenue as consumers and businesses adopt AI in every facet of American life.

If those assumptions about technology and demand prove wrong, these deals to clinch future capacity could become a monstrous burden for the tech companies and their investors.

Meta’s gigantic “Hyperion” data-center project in Louisiana, which is the size of about 1,700 football fields, helps explain how big obligations wind up off tech companies’ balance sheets.

Meta initially agreed to lease Hyperion for a four-year term starting in 2029, with options to renew for up to 20 years. It guaranteed that it would make bondholders whole if it doesn’t stay the entire two decades. The company doesn’t think payments under that guarantee are probable, so it hasn’t recorded any liability on its balance sheet.

In accordance with accounting rules, Meta’s Hyperion lease obligations will remain off balance sheet until it starts paying rent. It said its aggregate initial lease commitment is about $12.3 billion. Meta disclosed $347 billion in total obligations for leases that haven’t kicked in yet, including for Hyperion, as of June.

Across the companies the Journal analyzed, promises of payments under these uncommenced leases totaled $1.2 trillion in off-balance–sheet obligations, or about four times more than what was disclosed a year earlier. In addition to Meta, the Journal reviewed commitments for Alphabet, Amazon.com, Microsoft, Oracle, Nvidia, Broadcom, SpaceX and Advanced Micro Devices.

Data centers get stuffed with a lot of hardware, including the Nvidia chips that are used to train and run models and memory chips that store information. To buy all that, companies sign long-term contractual agreements well in advance to lock in production from their suppliers.

Those and other purchase obligations at the companies the Journal examined stand at a whopping $1.9 trillion. Under accounting rules, purchase commitments typically remain off balance sheet until a product or service is delivered. [...]

For the more anxious set on Wall Street, it is a worrying sign that some tech companies that once seemed to have fortress balance sheets have needed to tap the capital markets frequently.

Alphabet and Amazon recently posted results showing negative free cash flow, meaning their capital spending exceeded the cash they brought in from operating their businesses.

And that is before considering the implications of trillions in off-balance–sheet commitments. Whether or not the revenues ever arrive, purchase commitments and signed leases can’t be canceled, for the most part.

If things go wrong, tech companies will be paying an expensive tab for infrastructure that they can’t profitably use. These obligations could also lead increasingly indebted companies to have to borrow even more.

by Peter Rudegeair and Peter Santilli, Wall Street Journal | Read more:
Image: WSJ

Wednesday, August 12, 2026

Why Progressives Are Backing Off “Woke 1”

Earlier this week, Rep. Alexandria Ocasio-Cortez (D-NY) flashed her political dexterity on ABC’s This Week when she was asked about the policies of Peak Woke — call it Woke 1, that stretch around the 2020 pandemic when the American liberal machine seemed to be at its political and cultural height. “I have a local city councilman who has this saying,” the Bronx congresswoman said. “Woke 1 was crazy.”

She was quoting a tweet from Chi Ossé, a New York City Council member from Brooklyn, but the point underneath it was the real tell. The Democrats’ emerging left — fresh off primary victories in New York, Colorado, and now Michigan — has developed a playbook for handling any unpopular stance, or tweet, from the early 2020s, whether it’s pandemic masking, defunding the police, or the broader vocabulary of that summer.

The new plan: Brush it off. Then refocus on the present, and on the policies that will actually make people’s lives better.

So far, the strategy is working. Zohran Mamdani walked back his support for defunding the police during his 2025 mayoral campaign and is now the mayor of New York. Darializa Avila Chevalier, the DSA-backed congressional candidate in New York’s 13th District, had a stack of deleted posts that CNN’s KFile resurfaced this June — including one that read “all deportations are wrong.” When I asked her about them, she didn’t disavow a word. She won her primary, albeit in a deep-blue district, anyway. Wisconsin gubernatorial candidate Francesca Hong had a bad case of the woke mind virus in 2020 — the “cancel Thanksgiving” kind — and she’s still in the hunt in this week’s Democratic primary, where her candidacy has been far more focused on issues like data centers and education funding.

Electorally, at least, Ossé has been proven right. The pandemic was wild, and primary voters seem to be extending progressives a grace period of sorts — a chance to reintroduce themselves after workshopping their ideas over the last five years. And while Republicans are still hopeful they can relitigate some of these fights in a general election, they’ve struggled to make them stick in high-profile races since 2024.

But these are also the issues — race, gender, sexuality, affirmative action, religion, cancel culture — where the simplest story gets repeated out of ease. The national media’s distance from everyday people, and especially from working-class people of color, is felt most in how it covers culture. Catch-all terms like “wokeness” flatten real differences across the electorate and quietly impose a conservative frame on genuine arguments about equity and inclusion.

That was one of my biggest takeaways from that 2020 summer: our collective discomfort with all of these issues in the first place. I watched that year up close, traveling with Democratic and Republican candidates through the pandemic and the racial-justice protests, and there was never a clean “Two Americas” moment — at least not the left-versus-right one everyone expected. The real gap was between Masked America — coastal, professional, absolutely consumed by identity politics — and everyone else. Among ordinary voters of all parties, and especially minority voters, there was significantly more skepticism from the start toward the highly-specific “woke 1” litmus tests now being renounced by their leading proponents. At George Floyd’s memorial in Houston and in the city where he died, there was no consensus on defunding the police, only a shared agreement about racial injustice and a deep distrust of institutions. In south Chicago, people were already skeptical that the sudden flood of corporate money into racial justice would outlast the moment — even while the checks were still being written.

So no, I don’t love the tidy language of Woke 1 (then) versus Woke 2 (now). But here are a few lessons from that summer — and from our recent reporting on America, Actually — that I think apply now, especially to the progressives stepping back from things they used to say.

1. Woke isn’t dead

The broad ideals — diversity, equity, a basic sense of fairness — still matter to Democratic voters. Criminal justice reform still has real purchase with the base. Minority and women candidates have kept winning since 2024, and even ideas like reparations still resonate with big chunks of the Democratic electorate. What’s changed isn’t who these candidates are — it’s what they choose to lead with.

What has fallen out of favor is a way of talking. Let’s say it’s the nonprofit register, or the Ford Foundation cadence — the language that lived in mission statements and land acknowledgments, or that made sweeping generalizations about a specific group — white women, cis men, you name it. David Axelrod, describing Wisconsin’s Francesca Hong, gave it this label: a “font of zany liberal, faculty lounge exotica.” Yes, that’s the part that’s gotten less popular. The vocabulary changed more than the values.

2. The left rewrote its message after 2024

The clearest change is the affordability pivot. When Abdul El-Sayed ran for Michigan governor in 2018, the animating cause was climate and a Green New Deal. Today the center of gravity has moved to economic populism, affordability, and a rethinking of the US relationship with Israel. Climate didn’t vanish, but it’s no longer the headline. And “representation” was, at its core, a white-collar frame — it spoke loudest to people already inside the room.

Affordability speaks to everyone. When I sat down with El-Sayed on the show, this is how he described what Michigan voters were actually asking:
“Who poses the biggest pushback to a system of politics that has been bought off in ways that leave me unable to afford my groceries, unable to afford a home, unable to look at my kid’s school and believe that that’s a good place for them, unable to get healthcare when I need it — and then sending my money abroad, telling me that somehow that’s in my best interest?”
That’s not the language of 2020. It’s economic, universal, and it doesn’t ask anyone to first pass a vocabulary test.

It helps to remember that Sen. Bernie Sanders (I-VT) was never really “woke.” He talked about class more than culture, the many against the money. And it repeatedly got him into trouble, leading him to adopt more of the left’s rhetoric and positions on identity and race after 2016. In 2028, the sweet spot is probably the middle, and that’s what the next generation of progressives — AOC, Mamdani, El-Sayed — can do better than their Burlington forefather. They’re fluent in blending the class critique and the social one without missing a beat.

“This is about the many versus the money,” El-Sayed told me. “If you support a politics of the UAW, of working people everywhere, of teachers, of nurses, of working families, of Bernie Sanders and AOC, of people who want to break the chokehold of corporations and special interests on our politics — this is that race.”

by Astead Herndon, Vox |  Read more:
Image: Mario Tama/Getty Images
[ed. How convenient, a simple rebranding to create distance from idiodic policies that should never have gained traction in the first place. I'm an Independent because I don't trust Democrats to always do the right thing, although I do believe they're orders of magnitude better than Republicans, who nearly always gravitate to the wrong thing - or more reliably, the self-interested side of the spectrum, everyone else be damned. What's lacking in both parties are adults in the room that will put their foot down and cut off the wackier elements of both parties, who by virtue of their wackiness garner the most attention and water down the party's main themes (while providing endless ammunition to their opponents). Are there any adults left in either party with enough gravitas to say enough is enough? Hard to say, since we're talking about politics, where everything is poll-driven and transactionable. See also: Are the Democrats going to save us this time? (Noahpinion).]

Tuesday, August 11, 2026

How to Get Chatbots to Give Accurate Financial Advice

Finding good financial advice can be stressful – and expensive. That’s one reason why chatbots have become an increasingly popular and free alternative.

But using artificial intelligence to answer your pressing money questions also carries hidden dangers. I’m a finance professor who has been closely watching the spread of AI into personal finance, and I recently warned that AI is riskiest when it sounds most confident. I advised readers to bring in a human professional for high-stakes financial decisions. [...]

For people who can’t afford ongoing advice, AI is genuinely useful for budgeting, paying down debt and low-cost investing.

The skill lies in using AI well. Here are some simple guidelines to get accurate and actionable answers when you engage with a chatbot: [...]

Five habits that make AI safer

Once the list of questions is set, here are some precautions to take once you engage with a chatbot.

Make it ask you questions first. Open with, “Before you advise me, ask me the questions a good financial planner would ask.” Generic answers come from under-specified questions, and you learn which details will actually produce a more useful outcome.

Ask it to argue against itself. After any recommendation, reply: “Give me the strongest case against this, and the situations where it would be wrong for me.” If it can’t engage in response, that’s a sign the bot is entering a more dangerous mode. This one precaution does more than any other to signal for you to be careful.

Make it show its assumptions. If the bot projects that your savings will grow to an impressive number, ask what assumption it made and what would change it. You’ll learn that it assumes steady returns every year, no missed contributions and no fees. That means the projection is just information, not a promise.

Verify the facts. Contribution limits, tax brackets and deadlines all change, and this is exactly where AI can be subtly out of date. Check the IRS or the Social Security Administration directly. If one number drives your decision, don’t take it on a chatbot’s word.

Never share identifying details. Don’t offer account information, Social Security numbers or logins. Describe your situation in general terms. Good advice doesn’t require handing over data that can be used against you.

by Pawan Jain, The Conversation |  Read more:
Image: Badhan Ganesh on Unsplash, CC BY

Thursday, August 6, 2026

Americans Are Already Paying Dearly for the National Debt

Fiscal hawks like to drum up interest in the national debt by making the astronomical numbers more tangible. The United States owes $31.6 trillion to public creditors, more than $290,000 for each household. You could spend $1 million every day for almost 86,000 years before having to borrow more. But no one really cares. Talking about how many times all of the dollars laid end to end would go to the moon and back (6,000, as it happens) is just not going to get people to think differently about the national debt.

What should matter is that the consequences of this debt are not off in the future, but already here. The government’s deficits have saddled many American families with higher costs, largely from rising interest rates. The Budget Lab, the policy research center at Yale where I am the executive director, recently estimated that congressional-spending decisions since 2015 have raised Treasury yields by almost a full percentage point, which affects what American households pay to borrow. For someone taking out a 30-year mortgage at last year’s median home price, this rise in long-term interest rates has increased their borrowing costs by about $2,500 a year, or roughly $76,000 over the life of the loan. (The Budget Lab has built a tool to help users calculate their own extra mortgage costs.)

The problem is not just for Americans who are lucky enough to buy a home. The bloated government budgets and waning federal revenues of the past decade are driving up costs across the board. Compared with a world in which these fiscal-policy changes did not take place, the annual borrowing costs on a typical auto loan are now up by about $120, and by about $770 on a typical small-business loan. Credit-card borrowing rates are also hovering near record highs.

Although affordability has become a watchword for politicians who understand that rising prices are hurting American families, lawmakers seem to have forgotten that reducing federal deficits would help bring down prices. In the 1990s, Congress and the White House prioritized bringing deficits down by both cutting spending and raising revenue—moves that lowered borrowing costs for American families by about 0.6 percentage points, according to Budget Lab calculations. But few lawmakers seem to be suggesting the spending cuts and tax increases necessary to lower costs now. [...]

Much of the big legislation of the past decade, such as the Tax Cuts and Jobs Act, pandemic stimulus bills, and the One Big Beautiful Bill Act, has grown the deficit. Lawmakers have passed some legislation to improve the fiscal outlook, such as the Fiscal Responsibility Act in 2023, which cut spending and clawed back unspent coronavirus-relief funds, but most federal policy has lately involved spending money that the country doesn’t quite have. This is hurting consumers, businesses, and the federal government.

The cost of the war in Iran, which the Pentagon put at $29 billion last month (other estimates are higher), will put slight upward pressure on interest rates (0.002 percentage points), according to our calculator. The One Big Beautiful Bill Act, which we estimate will raise the deficit by $2.4 trillion over the next decade (not including interest costs), will raise interest rates on a typical 30-year mortgage by 0.4 percentage points by the end of 2030—about $1,060 annually for a home bought at the 2024 median price with a 20 percent down payment—and by 1.5 percentage points by the end of 2055.

Most economists support deficit spending during temporary crises, such as a recession, or in cases where an investment can be expected to generate more government revenues in the future, such as funding for infrastructure. But the United States has been spending far more than it takes in for well over two decades.

The main remedies for these problems—higher taxes and spending cuts—are generally politically unpopular. Every budget fix will have its critics, but some options are more palatable than others. Better funding for the IRS, for example, could help close the “tax gap”—the amount of taxes legally owed that are not paid in a timely way—which the IRS estimated at about $700 billion a year in 2022. Other levers include raising the retirement age and reducing Social Security benefits for high earners, who also tend to live longer; reforming Medicare Advantage, a program that has been shown to allow private insurers to overcharge the federal government; and removing the tax exemption on employer-provided health insurance, so that these benefits can be taxed as income. The Congressional Budget Office regularly publishes policies that could help close the deficit, and Americans need to decide what we’re willing to pay for and what we’re not.

A big challenge in making these hard choices is that the costs and benefits are asymmetrically understood: Whereas the costs of deficits are diffuse, the costs of policies that close the deficit are acutely clear only to those affected. For example, the Budget Lab has estimated that closing the carried-interest loophole could raise more than $100 billion in federal revenues over 10 years, which would help lower mortgage rates by 0.0064 percentage points. But this collective benefit is too slight for most people to know or care about it. The few people who benefit from this tax break, however, in industries such as private equity and venture capital, very much do care, so they are far more likely to push hard to keep it than the millions of affected Americans are to push to end it.

Politicians respond to electoral consequences. Right now there is nothing stopping them from doling out tax cuts and spending promises while also driving up interest rates. Voters may complain that their lives are becoming unaffordable, but hardly anyone seems to appreciate that federal deficits are partly to blame. If we want to see lawmakers actually address this problem, economists need to do a better job explaining the stakes. This means that instead of talking about the fact that our national debt could fill all 32 NFL stadiums with two tiers of construction pallets filled with $100 bills, we should be talking about how deficit spending is making it harder to pay our own bills.

by Martha Gimbel, The Atlantic | Read more:
Image: The Atlantic. Source: Getty
[ed. See also: America is Heading for a Debtpocalypse (Noahpinion):]
***
As of 2026, we’re in double trouble. Our national debt is back up above 100% of GDP — similar to what it was right after WW2 (and much higher than in 1990). But now the interest rates our government has to pay on its debt are almost twice as high as they were after WW2: [...]

But things are worse under Trump than they were under Biden, for three reasons.

First, this is a very large annual deficit, and it’s all being borrowed at the new, higher interest rates. In addition, during Biden’s first two years in office, inflation eroded the debt. Inflation is back down to a fairly low-ish level now, meaning the debt isn’t getting eroded. And finally, interest rates have now been high for long enough that the debt Trump borrowed in his first term to pay for Covid relief is now being rolled over at higher rates.

So right now, the national debt continues to explode, because the government is borrowing money just to pay the interest on the money it borrowed before. This increased debt naturally results in even greater interest costs, forcing the government to borrow even more to fund those interest payments. And so on. Interest payments and debt just go to the moon.
***
[ed. Let that sink in - we're paying interest on loans we've taken out to pay interest on the national debt. Also: Federal Debt 101; and Going For Broke (DS). And this: The Fiscal Crisis Facing American Cities (Urban Proxima):]
***
As the cost of servicing the debt increases, Congress must borrow more, raise additional revenue, or devote a smaller share of the federal budget to everything else. Whichever path it chooses, the federal government will have less room to maintain the commitments on which American cities have come to depend. [...]

Federal money flows to cities in three flavors: direct transfers, indirect transfers, and what we call fiscal dark matter. Direct transfers are exactly what they sound like — money sent directly from the federal government to various localities. These include funds disbursed through programs like the Community Development Block Grant (CDBG), which supports things like public infrastructure and neighborhood services. In 2022, direct transfers like the CDBG totaled $146.3 billion. That’s significant, but actually the smallest of the three categories.

Less visible are the indirect transfers. These monies are initially awarded to state governments, which then allocate funds to municipal-level programs and services in accordance with state prerogatives. The cleanest example is probably K-12 education, which receives federal Title I dollars to pay for teachers and programs, but federal highway dollars work essentially the same way. All told, in 2022, the federal government handed down $1.1 trillion to state governments. That amounted to 36% of overall state revenue for that year and, depending on the individual state, ranged from roughly 22% to 50% of state revenue. How much of that ultimately flowed down to cities is hard to say, which is itself a problem: it’s difficult to even establish how exposed local governments are to a pullback in federal support of state budgets.

The third category – our fiscal dark matter – is all the federal money spent into local communities that never shows up in a local budget. This includes housing subsidies like the Low-Income Housing Tax Credit (LIHTC) and Section 8. It also includes food support programs like SNAP and even some direct funding for local food banks.

Rightfully or not, when the flow of federal money in this category starts to dry up, the resulting problems will fall squarely on the mayor’s desk. After all, the median voter is never going to see increasing numbers of homeless encampments and think to blame the head of HUD. [...]

The “eds and meds” economies that anchor cities like Pittsburgh, Cleveland, and Baltimore depend heavily on Medicaid reimbursements and federal research grants to sustain the hospitals and universities that rank among their largest employers. Cuts there could precipitate layoffs in the institutions that have been holding together post-industrial downtowns for 30 years. And therein lies the second part of the dark matter problem. Federal money doesn’t just fund services and pay for infrastructure. In some places, it also props up major employers who anchor the entire local labor market. [ed. And the integral supply chain business that support those services.]

Wednesday, August 5, 2026

Sunday, August 2, 2026

Feds Implement Temporary Water Sharing Agreement in Western States

Arizona, California and Nevada will be required to curb their use of the water from the Colorado River by about 20 percent over the next two years — and could ultimately face even larger cuts — according to three officials familiar with negotiations over a long-awaited federal plan to rescue the depleted river.

The plan, part of which the Bureau of Reclamation is expected to describe in an Environmental Impact Statement on Friday, comes at a time of escalating crisis for the Colorado, a crucial water source for seven states, 30 Native tribes and a swath of northwestern Mexico. But experts say it will not be sufficient to resolve a political standoff among the river’s many users or prevent the beleaguered waterway from teetering toward collapse.

The cuts proposed for the next two years resemble what the three states offered in a proposal this spring, and represent the first phase of a broader 10-year framework for operating the river’s dams and reservoirs, according to the officials, who spoke on the condition of anonymity to discuss ongoing negotiations.

That framework is expected to call for operating plans to be developed every two years and outline a wide range of possible measures those plans could include — including reducing the amount of water released to the Lower Basin by as much as 40 percent.

The framework is not expected to consider mandatory cuts to water use from the four states in the upper part of the basin: Colorado, New Mexico, Utah and Wyoming. Arizona, California and Nevada make up the Lower Basin. [...]

The current operating rules, which expire at the end of September, have not prevented chronic overuse of the river amid a decades-long drought worsened by climate change.

After a historically meager winter snowfall and a scorching spring, the amount of water flowing into the river this year is less than a quarter of average annual demand, and levels in its major reservoirs have dropped to record lows. Scientists warn that one or two more dry years could crash the entire system, disrupting hydropower production, drinking water supplies and irrigation for some 5 million acres of farmland. [...]

The likely operating plan for 2027 and 2028, based on a May proposal from the Lower Basin states, is projected to save about 3.2 million acre feet of water — enough to fill roughly 1.5 million Olympic swimming pools. The plan will require significant “belt tightening,” particularly in Arizona, according to Sarah Porter, director of the Kyl Center for Water Policy at Arizona State University, but states have indicated they can tolerate the reductions.

Yet those measures are only half of what studies suggest is needed to bring water demand in line with the dwindling supply, Porter cautioned, increasing the likelihood of even steeper cuts down the road. [...]

The 330-mile system of canals and aqueducts, which supplies water to the most populated parts of Arizona, is poised to see the biggest cut in its history under the bureau’s operating plan for the next two years. If the agency chooses to implement some of the deeper reductions considered in the 10-year framework, CAP’s entire water allocation could be wiped out. [...]

Fraught negotiations

Experts say the rising tensions on the river result from a chaotic combination of bad weather, poor planning, intransigent state officials and federal missteps under the Trump and Biden administrations.

At the heart of the conflict is an impasse between the Upper and Lower Basin states over who should shoulder the burden of necessary cuts.

In the Upper Basin, home to the snowcapped mountains and winding tributaries that feed the river, there are few reservoirs to provide long-term water storage, leaving users reliant on natural flows. That means the Upper Basin takes an automatic cut during dry years, officials argue. They say responsibility for restoring water to Lakes Powell and Mead should fall on the Lower Basin states that use them.

Yet about three-quarters of the people who depend on the Colorado live in the Lower Basin. The region is also home to major cities and sprawling farms that provide most of the nation’s winter vegetable supply. Officials from these states say they have already curbed their water consumption by millions of acre feet in recent years. Overuse of the river is universal, they argue, and so too is responsibility for saving it.

The situation is complicated by the arcane legal framework governing the river, which prioritizes users chronologically. Without agreements among the states, major cuts would fall entirely on junior users, including huge cities such as Phoenix and Tucson, before more senior rights-holders such as the farmers in California’s Imperial Valley see any reductions.
Last summer, it looked like states might agree on a new method of apportioning the river based on actual flow, rather than historical averages and legal agreements. But those negotiations broke down over familiar disagreements about who should be subjected to mandatory cuts. [...]

A vanishing river

Brad Udall, a climate scientist at Colorado State University’s Colorado Water Center, describes the tensions over the river as a “big collision of 19th-century water law, 20th-century infrastructure and 21st-century climate change and population growth.”

The Colorado has almost never contained enough water to satisfy everyone who has legal rights to it, Udall said, and human-caused warming has made the situation even worse. Since 2000, high temperatures and shifting rainfall patterns linked to climate change have diminished the amount of water flowing through the river by about 20 percent, compared to the 20th-century average.

The deficits have forced repeated negotiations over how to manage shortages. Past deals have helped curb consumption somewhat, but they were never stringent enough to reverse the inexorable decline of reservoirs that are intended to provide a buffer during bad years.

Lake Mead, the site of the Hoover Dam, is mere inches from its lowest level on record. A few hundred miles upstream, Lake Powell is approaching the point at which water can no longer flow through the turbines of the Glen Canyon Dam. That raises the risk of a phenomenon called cavitation, in which air bubbles form then implode in fast-moving water, releasing energy that can damage the dam itself.

“The reservoirs are depleted so low they’re really at the end of their capability,” Castle said. “We’re in such a precarious situation.”

by Sarah Caplan, Washington Post |  Read more:
Image: Caroline Brehman/Reuters
[ed. The U.S. Bureau of Reclamation on Friday unveiled the framework that will guide operations on the Colorado River through 2036. See also: Lake Powell's Dying Days (CCG):]
***
The L.A. Times’ Ian James reported that Trump’s Interior Department would accept a proposal submitted by California, Arizona and Nevada — the Lower Basin states — to slash their water use by 12%, 31% and 28%, respectively, through 2028. They’ll receive $350 million from Biden’s Inflation Reduction Act to support water conservation.

The Upper Basin states — Colorado, Utah, New Mexico and Wyoming — will get $100 million in conservation funding. But unlike their downstream neighbors, they won’t face mandatory water cuts. However much water they end up saving, that will be good enough. [...]

If Powell’s water levels sink much lower, water won’t be able to pass through the dam’s hydropower turbines, which generate cheap electricity for communities across the West. That wouldn’t be a “dead pool” situation; water could still flow downstream to the Grand Canyon and Lake Mead through bypass tubes lower in the dam. But the bypass tubes are surprisingly frail and could break with sustained use.

Translation: We are frighteningly close to “de facto dead pool.” That’s why the Trump administration is ordering everyone to use less water.

Well, not everyone. California, Arizona and Nevada are willing to cut back dramatically, and federal officials seem happy to make them do it. The Upper Basin states — the ones upstream of Lake Powell — say they shouldn’t have to commit to mandatory reductions, in part because they already consume a lot less.

In a New York Times opinion piece earlier this year, I argued that the Upper Basin states need to do more. Podmore agreed.

“It’s a tricky situation, because the Lower Basin has always used more water, and that’s a convenient argument for the Upper Basin,” he said. “But also, there’s more people in the Lower Basin. And the most productive agricultural land that’s irrigated with Colorado River water is located in the Lower Basin.”

“Even with the cuts that the Lower Basin has offered, we still have a long way to go to balance the water budget,” he added. “Everyone needs to pitch in.”

[ed. But not everyone is agreeing to pitch in: California’s Biggest AI Data Center Is Suing for Colorado River Water (Yahoo News):]
***
The developer behind California's biggest planned AI data center publicly swore it would never touch Colorado River water. It would run on recycled wastewater — clean, virtuous, zero environmental impact. That pledge held right up until the cities of Imperial and El Centro said no thanks. Now Imperial Valley Computer Manufacturing (IVCM) has sued the Imperial Irrigation District (IID) for access to the very river it promised to leave alone. The facility would sit in a desert valley where 180,000 people share exactly one freshwater source.

The Farm-to-Cloud Gambit

IVCM's legal strategy treats 160 acres of fallowed farmland as a water entitlement for a nearly million-square-foot AI campus.

The developer's playbook relies on a tactic called "buy and dry" — purchasing irrigated farmland, retiring it from production, then claiming its water allocation for industrial use.

Will Larry Ellison Be the Face of the A.I. Bubble?

[ed. Don't miss this one. It's got everything (and could easily be a Pulitzer contender).]

On Jan. 21, 2025 — the first full day of the second Trump administration — Larry Ellison woke up in his 33-bedroom, 34-bathroom oceanfront mansion in Florida, got into his Gulfstream jet and headed up to Washington. Ellison, who was 80 and worth in the neighborhood of $200 billion, had an appointment at the White House. He didn’t bother to take a driver’s license — he needed to call someone on the president’s staff to vouch for him at the gate — but there he was, at 2 p.m., standing beside Donald Trump in the Roosevelt Room as the president announced “the largest A.I. infrastructure project by far in history” and told the world that his friend Larry Ellison was just the man to get it done. “He’s sort of C.E.O. of everything,” Trump said. “He’s an amazing man and an amazing businessperson.”

Ellison began by thanking Trump. “We certainly couldn’t do this without you,” he said. “It would simply be impossible.” He then proceeded to sketch out the ambitious plan. Ellison’s database software and cloud computing company, Oracle, and its partners — most prominently OpenAI — were going to invest as much as $500 billion over the next four years into a group of sprawling data centers, 500,000 square feet each, that would produce 10 gigawatts of computing power, using enough energy to power as many as 10 million homes. It was called Project Stargate, after the 1994 sci-fi movie in which Kurt Russell steps through a wormhole and finds himself inside a pyramid on an alien planet. This Stargate would be a portal leading humanity from the postindustrial era to the artificial-intelligence age. [...]

For Ellison, it was the capstone of a mad two-year scramble to transform Oracle into an A.I. juggernaut. The effort began in late 2022 when the launch of ChatGPT stunned the world and set in motion a race to master and control the most transformative new technology since the birth of the internet. Ellison, a founding father of Silicon Valley and the last of his generation still in the game, was desperate to avoid getting left behind. He’d moved quickly and aggressively — some might even say recklessly — to turn Oracle into a “hyperscaler,” one of the handful of companies providing the critical infrastructure that would power the A.I. boom. [...]

ChatGPT landed very differently in Washington than it did in Silicon Valley, setting off a scramble of its own inside the Biden administration to regulate the development of A.I. To oversee his A.I. policy, Biden turned to a veteran Democratic policy adviser, Bruce Reed, who believed that the administration needed to be proactive. A year after ChatGPT’s debut, in late 2023, Biden signed a comprehensive executive order on A.I., seeking to define the government’s role in the future of this new technology.

For the Biden administration, artificial intelligence was by no means just a domestic economic issue. Countries around the world were all racing to develop their own A.I. infrastructure and technology, and global power and influence would flow to whoever got there first. From this perspective, A.I. data centers were less businesses than geopolitical assets.

The administration was especially concerned about the A.I. ambitions of China and the Persian Gulf, given the powerful role artificial intelligence was likely to play in reshaping the information ecosystem. [...]

The administration’s concerns and Ellison’s ambitions were on a collision course. China and the Gulf were both critical to Ellison’s A.I. plans. Oracle already had a lot of contracts around the Gulf, and it also had a strong business relationship with one of China’s most important A.I. companies, ByteDance. Oracle was the U.S. cloud provider for the U.S. division of ByteDance’s TikTok, storing and securing the data of the app’s 100 million American users. But with ByteDance itself now pivoting into generative A.I., they had the opportunity to do more business together. In the summer of 2024, Oracle started working on a $6.5 billion deal to build a large data center complex in Malaysia, from which it could convey computing power to ByteDance and other foreign companies through opaque leasing deals.

It would be perfectly legal — but under the Biden administration maybe not for long. By that point, national security officials were growing increasingly concerned about China and the Gulf’s A.I. ambitions and were discussing ways to gain more control over them. The administration was especially worried about the role Oracle might play in fueling these ambitions. They knew that Ellison was trying to scale up the company’s A.I. infrastructure quickly and that it was badly in need of cash, which meant that it might be more tempted to make deals that the administration didn’t think were in America’s best interests. [...]

In early 2024, the administration started working with Congress on a bipartisan bill — the Protecting Americans’ Data From Foreign Adversary Controlled Applications Act — that would force ByteDance to divest its U.S. TikTok operations. Biden signed the bill into law in April 2024, setting a deadline of Jan. 19, 2025, for a sale. If ByteDance failed to meet the deadline, the app would be shut down in the United States.

At the same time, the administration was preparing to shore up its efforts to restrict China’s access to American computing power and to exert more control over the Gulf’s. In late 2024, it circulated the draft of a plan to require hyperscalers to go through a licensing process to operate overseas and to keep 50 percent of their computing power in America.

All of the hyperscalers were looking to build overseas, but Oracle had the most to lose: Its global plans were the most ambitious, at least relative to its size. The company publicly and aggressively opposed the Biden plan. Its top policy executive in Washington, Ken Glueck, called it “one of the most destructive” moves ever taken against the tech industry, arguing that the best way to solidify America’s lead in the artificial intelligence race was for U.S. companies to build and control as much of the world’s A.I. infrastructure as possible.

Biden signed off on the new policy in the final days of his presidency. It was scheduled to go into effect in May 2025. If enacted, it could force Oracle to scale back its ambitions in Malaysia and the Gulf. Ellison’s plan to transform Oracle was in trouble. But a new president was on his way to Washington.

‘The Tsunami’

Relief came almost immediately. Hours after his inauguration in January 2025, Trump sat down at the Resolute Desk and began signing executive orders aimed at dismantling Biden’s A.I. policies. He also signed an order directing his attorney general to hold off on enforcing the congressionally mandated TikTok ban for 75 days. And then, of course, came the Project Stargate announcement with Ellison and Altman.

Trump turned to a very different group of people to shape his new administration’s approach to artificial intelligence. He named as his A.I. and cryptocurrency czar David Sacks, a Silicon Valley venture capitalist who had raised many millions for the Trump campaign and, according to a New York Times investigation, was personally invested in at least 449 companies with ties to artificial intelligence. Sacks, who has denied any conflict of interest, believed that when it came to A.I., the government’s job was to get out of the way.

The National Security Council’s technology and national security division had played a key role in shaping America’s A.I. policy in the Biden years. Trump initially appointed David Feith — who had serious concerns about China’s ability to remotely access computing power through Malaysia and other Southeast Asian nations — to run it. But in April, he fired Feith and a few other China hawks and then eliminated the entire directorate. [...]

Trump saw another benefit to withdrawing the Biden plan: The Gulf states were adamantly opposed to it. They needed U.S. computing power to build out their own A.I. infrastructures and had something to offer in return. Their sovereign wealth funds were sitting on trillions of dollars that they were ready to invest in all sorts of American companies, including some connected to the Trump family.

Two weeks before the Biden policy was scheduled to go into effect, Zach Witkoff — son of the Trump adviser Steven Witkoff and chief executive of the Trump family’s cryptocurrency firm World Liberty Financial — made an announcement at a conference in Dubai: The Emiratis would use $2 billion of the firm’s brand-new stablecoin for an investment in Binance, a crypto exchange. Less than two weeks later — 48 hours before the Biden restrictions would kick in — Trump rescinded the policy.

That same day, Trump landed in Saudi Arabia, the first stop on a three-day tour of the Gulf. He was joined in the United Arab Emirates by Altman to announce Stargate U.A.E., a multibillion-dollar initiative to build one of the world’s largest data centers outside Abu Dhabi. Oracle would be a partner, too.

With the Biden plan dead, Oracle was free to operate its data center complex in Malaysia as it saw fit. By the end of June, the facility was on track to become the second-biggest in the world. Oracle doesn’t release the names of its customers there, but by studying its output, an independent A.I. research firm, SemiAnalysis, determined that the facility was feeding most of its computing power to ByteDance. An analyst at the tech-focused think tank ChinaTalk, Aqib F. Zakaria, ran his own numbers and arrived at a startling conclusion: Oracle was providing a staggering 22.6 percent of China’s known A.I. computing power.

by Jonathan Mahler, Jim Rutenberg and Kirsten Grind, NY Times |  Read more:
Images: Louie Psihoyos; Scott Ball
[ed. Not to be redundant but this came out shortly after I'd posted about Oracle (and Larry Ellison) below in The Hater's Guide to Oracle (Part 2). It contains a treasure trove of new information and a road map to how business and politics intersect in Washington and around the world these days. Well worth a read.]

Saturday, August 1, 2026

The Hater’s Guide To Oracle (Part 2)

Oracle has one of the strongest mythologies in the tech industry. Ask a regular person and they’ll tell you that it’s “incredibly profitable” and “growing fast,” that it’s “unstoppable,” and that Larry Ellison has the mandate of heaven with regard to the continual sales of software and hardware related to databases and AI.

And those people are completely and utterly wrong.

The original title of this article was “Is Oracle Dying?” because I assume, when I took a deeper look, that there’d be some sort of debate, some sort of bull case for a decades-old quasi-hyperscaler run by one of the more nakedly-evil CEOs in the history of tech. I assumed — incorrectly, I might add — that Oracle as a business was doing fine other than the ridiculous commitments it made to support the whims of Sam Altman and OpenAI via deals that I believed (and still believe) will kill Oracle.

Except it turns out that Oracle has already been on a death spiral for the best part of a decade (if not longer) and has only survived this long by screwing its customers, taking on masses of debt, and — most importantly — more than $85 billion in acquisitions over the last 23 years. Pretty much every major product line outside of databases is a hodge-podge of other people’s innovation stapled together with a legendary contempt for the customer. These acquisitions (and continual price increases) are the only thing keeping the reaper from Oracle’s door other than margin-destroying GPUs. [...]

After April 2009’s $5.7 billion acquisition of Sun Microsystems, Oracle’s revenues barely kept pace with inflation until December 2021’s $28.3 billion acquisition of Cerner allowed it to create Oracle Health, adding about $6 billion in annual revenue that had 40% lower margins (about 21.7%) than Oracle’s other businesses, though Oracle immediately started closing offices and brutal layoffs to try and bring them up.

And as I mentioned above, Oracle’s other plan was to sink a little over $99 billion in capital expenditures since the middle of calendar year 2020 into AI GPUs. [...]

Oracle is a decades-long mission to keep reapplying lipstick to a pig. Billions of dollars of acquisitions have, for the most part, only succeeded in keeping the company’s revenue growth from going negative, and as noted by forensic accountant Howard M. Schilit, this is one of the most well-documented cases of accounting shenanigans being used to cover up that a business is in decline.

Today’s newsletter is a sequel to the Hater’s Guide To Oracle, where I told the sordid tale of how Larry Ellison grew a massive, lucrative business out of a database business that one reporter once told me was a “law firm with a database company attached,” an Enterprise Resource Planning (ERP) product that competes with SAP to create the most-annoying way to run a large company, and a business built around licensing Java that exists mostly to email people and say “you need to pay us for Java or we’ll sue you.”

Then, as I’ve mentioned, there’s Oracle’s cloud infrastructure business, a decade-old also-ran that was meant to compete with Microsoft Azure and Amazon Web Services, but only managed to catch up following the advent of AI GPUs and a movement where all it took to party was buying billions of GPUs and saying “gosh darn, we love AI.”

I originally started drafting this as a much tamer piece where I’d ask whether Oracle was dying, but as my editor and I started digging into the research, it became obvious that not only is Oracle dying, it’s been dying for years, kept alive through decades of acquisitions and a desperate and dangerous commitment to generative AI.

And AI, I believe, will be what eventually kills Oracle dead. [...]

With revenue plateauing and customers in revolt, Oracle’s future already looked murky, but with the power of AI — and $95 billion in FY2027 capex — it’s becoming increasingly clear that this may be Larry Ellison’s last dance with Silicon Valley.

by Ed Zitron, Where's Your Ed At |  Read more:
Image: Larry Ellison, Bloomberg/Getty
[ed. Larry Ellison. One of the most hated personalities in tech (and unfortunately, owner of my beloved island of Lanai, in Hawaii). Update: What a coincidence. There's quite a story in the NY Times that just came out about Ellison being the face of the AI bubble. See also: The Hater's Guide to Oracle (Zitron); Ellison Empire Beseiged On All Fronts (NC); and, this excellent series The Oracle Files by Drey Dossier on YouTube. (For example, this one: How Larry Ellison and Gulf Money Just Bought Your News):]
***
Warner Brothers Discovery shareholders are getting screwed on this new Paramount deal. Okay. And I would like to get into exactly how before they vote on Thursday, the largest media merger in American history is going to a shareholder vote. A merger worth in the ballpark of $111 billion in case you were wondering.

Which means that Warner Brothers, you know, the big conglomerate that owns CNN and HBO, is potentially getting folded into another conglomerate Paramount Pictures, which is the company that owns CBS, MTV, Showtime, and Nickelodeon. And the shareholder vote is April 23rd, this up coming Thursday.

And last Thursday afternoon, which is one week before the vote, Warner Brothers Discovery filed a 14 page correction to the document that shareholders are voting o n.

Now, this is kind of a big deal because this is a 14page addendum to the biggest media merger in American history. And this was filed on Thursday of last week, 4 days ago at this point. 

Now, public companies don't usually rewrite their own proxy statements a week before a shareholder vote, unless of course someone is forcing them to, which usually means that someone being one of their shareholders is suing them in order to do so. So, I checked to see if there were any lawsuits floating around out there, and what do you know? There is one. A shareholder named Donna Nikosia, apologies if I butchered that last name, filed a lawsuit on April 2nd saying that the original document left out a lot of information that shareholders needed in order to make an informed vote. 

And following Donna's lawsuit were 15 other shareholders who had sent letters more or less saying the same thing. And can we all just take a moment here and say thank you to Donna for filing what we all probably knew to be true in the back seconds of our heads that there is information being left out that you need in order to make an informed decision this upcoming Thursday. Now up top I just want to say that I am not a Warner Brothers Discovery shareholder. I have never owned a share of Warner Brothers Discovery or Paramount Pictures. I am just thanking Donna as a media consumer.

All right, and somebody who works within the media ecosystem because I like to keep my media independent and this deserves a lot more scrutiny than it's getting. So WBD, Warner Brothers Discovery, told the court that this lawsuit had no merit and then two weeks later slightly added the information.

Anyways, so this move in business, I've learned, is how you smother out a lawsuit without ever having to say that we are wrong. Now, we're going to get into what was in this correction in a second here because oh boy, were they leaving information out? [...]

You know, I read that 14 page new filing this weekend and there are two companies in it that WBD is still trying very hard not to have to say out loud and is trying even harder, it seems, to smother this from any of the news outlets taking this to the other shareholders. And I think I figured out which ones they're talking about. 

[ed. And this: Why Iran's Blockade is an Oracle Story:]

Most people know Larry Ellison as the Oracle billionaire, which true, you also probably know that he is the largest private donor to the Israeli military in American history.

He's given over $26 million to the friends of the IDF since 2014, including a single $16.5 million donation in 2017. That is the largest gift in the organization's history. And that is the part we have discussed at length. But here is the part that a lot of people don't know. Ellison is not just the largest funer of the Israeli military. 

His company is the operational backbone of it. According to Open Intel, Oracle holds a 26-year contract to build and operate the IT infrastructure for the IDF's intelligence campus in Negv. For clarity, that is the facility that houses unit 8200, Israel's signals intelligence and cyber warfare division and one of the largest listening bases in the world. That is a 26-year relationship extending into the 2040s between a private American company and the intelligence apparatus of a foreign military. 

And that's just the intelligence side because Oracle also runs the Israeli Air Force entire logistics system, the supply chain that tracks his spare parts for F-35s and F-16s, aviation fuel and mutations inventory. Oracle hosts an AI battlefield management system called Fireweaver that coordinates sensors and weapons on the battlefield in real time, which means that Oracle software is making targeting decisions in the kill chain for Israeli Defense Forces.

Thursday, July 30, 2026

What Will More Intelligence Actually Do For Us?

In lots of sci-fi books, as soon as artificial superintelligence arrives, it bootstraps itself to even more godlike intelligence in an explosive “singularity” that rapidly transforms the entire physical universe. Lots of people, especially “AI safety” and “effective altruist” types, expected things to play out basically the same way in reality. But looking around, not much has changed since we entered the intelligence explosion. There’s a huge data center boom, and most people use AI on a daily basis, but we still live basically the same lives — driving to work or taking the train, sitting in front of a computer, scrolling on our phones, collecting a paycheck. People are staying in their jobs longer, but employment hasn’t been disrupted in a significant way:

A lot of people I know are surprised by this. Ruxandra Teslo writes:
Walking around the world today one might notice that it is weirdly unchanged…To many, this is surprising. Just the other day I was at a conference where someone remarked that if he could have seen today’s AI capabilities a few years ago, he would have been astonished — and would have assumed the world by now would look far more transformed, with much higher GDP growth.
Teslo blames bottlenecks — governance and other “frictions” — for the slow economic impact. But some others are advancing a more radical hypothesis — that intelligence itself is subject to diminishing returns.

One of these is Francois Chollet, an AI researcher who specializes in measuring AI’s capabilities. In a highly controversial series of tweets back in March, he conjectured that intelligence might be subject to diminishing returns:
One of the biggest misconceptions people have about intelligence is seeing it as some kind of unbounded scalar stat, like height. "Future AI will have 10,000 IQ", that sort of thing. Intelligence is a conversion ratio, with an optimality bound. Increasing intelligence is not so much like "making the tower taller", it's more like "making the ball rounder". At some point it's already pretty damn spherical and any improvement is marginal.
Now of course smart humans aren't quite at the optimal bound yet on an individual level, and machines will have many advantages besides intelligence -- mostly the removal of biological bottlenecks: greater processing speed, unlimited working memory, unlimited memory with perfect recall... but these are mostly things humans can also access through externalized cognitive tools.
In fact, this is a possibility I myself had raised in a post a year earlier:
It seems possible that humans are simply incredibly specialized in a few types of cognitive tasks — extracting patterns from sparse data, synthesizing various patterns into “intuition” and “judgement”, and communicating those patterns in language — and that we’ve basically approached the theoretical maximum in those narrow areas…That would explain why AI has gotten much better at things like math and coding and forecasting over the last year, but why the basic chatbot interface doesn’t seem much more “intelligent”. It would also explain why when you talk to Terence Tao about math, it’s like talking to a superhuman, but when you talk to him about where to get lunch or which movies are the best, he’ll just sound like a fairly smart normal dude. AI will eventually get better than Tao at math…but it may never get much better than the most thoughtful, eloquent humans at deciding where to get lunch or recommending movies. It may simply not be mathematically possible to get much better than we already are at that sort of thing.
Why would intelligence top out like this? Well, if we think of intelligence as the ability to extract information from data, then even an infinitely advanced model endowed with infinite compute will be limited by the fact that there’s a limited amount of information that can be extracted from the data.

For one thing, data itself is in limited supply. You can’t transform the world unless you can (in some generalized sense) understand it, and you can’t understand the world unless you can measure it, and our ability to measure the world is inherently limited and finite. [...]

So although we don’t know yet, it’s possible that humans were already hitting the point of diminishing returns with regards to individual cognitive capacity, and that superintelligent machines will never be as far beyond us as we are beyond dogs. But even if that’s true, I can think of at least three reasons why machine superintelligence could still deliver huge productivity gains. [...]

Distributed tacit knowledge

The German company Zeiss makes the best glass on the planet. If one of the mirrors that Zeiss makes for ASML’s EUV chipmaking machines were the size of Germany, the biggest bump on that mirror would be just one millimeter high. Only a few other companies — and maybe no other company on Earth — can match that. Zeiss’ mirrors also have a number of other amazing properties, like not distorting much due to temperature changes.

How does Zeiss make glass this good? No one knows — not even the people at Zeiss. If the technology were capable of being written down on a blueprint, China would have hacked Zeiss and stolen it, the way Huawei hacked Cisco and Nortel. If the technology were capable of being explained by a former Zeiss employee, or even several former Zeiss employees, China would have paid those people many millions of dollars to spill the beans.

Zeiss’ technology basically can’t be stolen, because it’s tacit and distributed. It consists of a vast number of little tricks and techniques that a huge number of individual employees use on a daily basis. These people don’t always even realize all those little things they’re doing that make the glass come out so good. And each employee knows a different set of tricks and techniques. The knowledge exists at the level of the organization itself, and is thus very hard to steal or recreate.

This is true of lots of corporate technology. A big part of the reason China can cut off the supply of rare earths to the rest of the world any time it wants to is that other countries aren’t very good at refining rare earths. Rare earths are difficult to separate from each other in solutions; it takes a ton of little chemistry tricks to do it cheaply at scale. Chinese refiners have spent four decades building up those little tricks and techniques; American or Japanese refiners won’t simply be able to replicate their efficiency overnight, and so it’ll continue to cost much more to produce rare earths outside China.

Except in the age of AI, this might change. Suppose American rare earth refiners give their employees a bunch of equipment to record everything they do — smart glasses, gloves, and so on — in addition to sensors distributed throughout their plants. AI will be able to synthesize all that information and very rapidly suggest small ways to improve the production process. Many of those little experiments will fail; others will succeed and will quickly be adopted, allowing another round of experimentation and improvement to begin very quickly. Crucially, AI’s ability to do this doesn’t depend on its raw intelligence — only on its ability to handle huge amounts of data very quickly.

In other words, in the age of AI, distributed tacit knowledge might not be nearly as big of a barrier to technological diffusion. This could improve economy-wide productivity, as lagging firms catch up to leading firms much more quickly. A more equal distribution of productivity would also make the economy more competitive, creating more surplus for consumers (though possibly reducing the incentive for firms to innovate, by making technology less excludable).

AI’s ability to quickly produce distributed tacit process knowledge might also supercharge productivity growth at the frontier. Imagine if any company could optimize any production process five times faster than today. The whole economy would speed up, as components got cheaper, turnaround times and product cycles got shorter, and scale-up got much faster.

And as with the previous example, improving the production of distributed tacit knowledge wouldn’t depend on AI’s raw intelligence. It would spring from AI’s ability to act like a computer — to interface directly with sensors, to handle lots of data, to perceive tiny details, and to do everything very very quickly.

by Noah Smith, Noahpinion |  Read more:
Image: Zeiss
[ed. Another thing I've wondered about: historians make a living unearthing little known facts and connecting dots from sources that are deeply buried in paper and microfiche respositories (and early data storage technologies - like 8 and 5 1/4 inch floppy disks). Millions of memos and correspondences that were once widely distributed and now sitting in dusty boxes or warehouses, archived somewhere. Items that could help significantly in understaning more about human judgement and decision-making. How much of this has been scraped for training? Very little, I'd presume.] 

Wednesday, July 29, 2026

We Asked Too Much of the American University

Our one remaining functional institution is going downhill.

The story of the rise and possible impending fall of the American university system is, in many ways, the story of modern America. It ties together the changes in our culture, our economy, and our politics since the mid-20th century. Understanding why universities came to be our most important and most functional institution, and why that model is now under threat, can help us understand how our country might look different going forward.

Let me try to tell a condensed version of that story.

The United States used to have a bunch of institutions that bound us together and forged us into a unified society. Many American communities were centered around churches; these facilitated networking, helped people find spouses, provided community services like day care and mutual aid, and homogenized values and culture at the local level. During the World Wars (and to a lesser extent, the Vietnam War), the military was very large and threw together Americans from various social classes. In the early postwar decades, corporations were also a unifying institution. Mass media provided us with shared cultural context. Even public transit put people of various backgrounds in close social contact on a daily basis.

In the half-century from around 1970 to 2020, those unifying institutions became much weaker. Church attendance slowly declined and then fell off a cliff in the 2010s. The military shrank into a small, professionalized volunteer force. Corporations ended their brief flirtation with lifetime employment, and outsourced many of their roles. Mass media fragmented in the age of the internet. Public transit dwindled in importance as Americans moved to the suburbs and drove.

As the country’s unifying institutions withered, one new institution attempted to step into the void: the American university. Universities were not new in the late 20th century, of course, but mass attendance certainly was. From 1970 to 2020, the share of Americans age 25-29 with a bachelor’s degree went from 16.4% to 39.6%. Around two out of three have completed some college.

College went from something that only the upper crust did, to something that most people were expected to do if they wanted to be economically successful in life. This expansion inadvertently but inevitably thrust a new role on American universities — that of a broad socially unifying institution. College was where people from a variety of backgrounds mingled and mixed — not just in classes, but in dorms, campus activities, and college towns.

College became the new church, the new military, and the neighborhood bowling alley all rolled into one. Instead of preachers homogenizing Americans’ values from the pulpit, university administrators taught college kids to value things like diversity, consent, and so on, and college students hashed out their own differences in millions of late-night dorm room discussions. Instead of meeting their first love in high school, Americans increasingly delayed sex until college; many of these relationships turned into marriages. Young people went into college as children and came out adults.

This was a heavy burden to bear for an institution that hadn’t been designed for it. But American universities were accustomed to taking on big new duties. Our universities began as essentially a copy of the British model — teaching-focused institutions designed to provide broad education and mentorship to the upper class. In the early 20th century they tacked on the German model — a research-focused lab apprenticeship system by which top scientists taught other top scientists and readied them for research jobs while also producing cutting-edge basic research.

This dual system enabled a remarkable form of cross-subsidization. Undergrad tuition payments — and state support for undergrad education, and undergrad alumni gifts — created a flood of money that paid for grad student stipends, lab facilities, professor salaries, and more. The federal government and companies also funded university research, of course, through grants and sponsorships. But undergrad money was a huge tailwind. And the prestige generated by successful and famous researchers helped universities charge undergrads more.

The hybrid of the old British and German models was naturally symbiotic, and it meant that even as America’s other institutions came under pressure, universities thrived and grew — especially once they used their prestige to attract high-paying and highly skilled foreign students from around the globe. Universities became the lynchpin of America’s research effort:

Source: Arora et al. (2019)

American universities had a lock on both the nation’s research output and on its production of human capital, and those roles were mutually reinforcing.

I suspect that their success at handling education and research at the same time probably gave American universities a lot of confidence about their ability to handle society and culture as well. Colleges spent more and more on dorms and “student services” and hired administrators (many of whom dealt with undergrad life) at an astonishing rate.1

But replacing churches, the military, and the neighborhood bowling alley proved harder than replacing the corporate lab had been. There was just one basic problem with college as America’s primary unifying institution, which is that not everyone can go to college.

First of all, college is difficult. To complete college courses, you need some degree of raw intelligence, but you also need work ethic and a certain degree of independence. As much as we might like to believe otherwise, not everyone in America has those traits, and we don’t yet know how to instill them in everyone. As a result, there’s a limit to how much you can expand college enrollment — and college completion — without loosening standards.

In fact, loosening standards is exactly what American colleges have done. Universities have necessarily become less and less selective over the years, as they have taken in a larger and larger fraction of the young American population. For a while, this led to lower college completion rates, but in the 1990s, more and more students began finishing their bachelors’ degrees. Why? Because colleges implemented grade inflation, making it easier to finish school without learning the material well. Here’s Denning et al. (2021):
We find that most of the increase in graduation rates can be explained by grade inflation, and that other factors such as changing student characteristics and institutional resources play little or no role. This is because GPA strongly predicts graduation and that GPAs have been rising since the 1990s. This finding holds in national survey data and in records from 9 large public universities. We also find that at a public liberal arts college, grades increased holding performance on identical exams fixed.
At some point, though, this process hits a wall. A large fraction of young Americans just isn’t prepared for college, even with grade inflation, and doesn’t end up going. College enrollment by recent high school graduates plateaued in the early 2000s and actually fell back to early 1990s levels during and after the pandemic:

Source: NCES via GPT

by Noah Smith, Noahpinion |  Read more:
Image: Ajay Suresh via Wikimedia Commons
[ed. Unfortunately, that's it before you hit the paywall. But you get the idea, and can claim the rest of the article for free once (by providing your email).]

Tuesday, July 28, 2026

There Are No Known Commodity Resources in Space That Could Be Sold on Earth

This blog is part of a series tackling common misconceptions in space journalism.

One common trope of space journalism these days concerns the mining of asteroids or the Moon, sometimes combined with environmental handwringing over the aesthetic destruction we may bring to these soulless dino-killing space rocks. Moon mining, we are told, is a gold rush about to happen. In the process, a few people will get super wealthy selling shovels or shiny metal of some kind, and hopefully a few big cities will get built in space. Indeed, space mining is sometimes seen as the “killer app” necessary to fund and motivate large scale human occupation of space.

Advocates of the industrialization of space usually envision a bootstrapping process, wherein one core product provides the profit margin necessary to build out infrastructure and, eventually, move most of Earth’s industry into space.


The question: Where is the space gold mine? While industrial processes add value at every step, space is often seen initially as a source of raw materials. Specifically, asteroids, the Moon, or Mars are seen as sites for future mines. These mines could produce anything from water to gold, Helium-3 to platinum. In this post, I will cover factors general to all material products before diving into specific examples.

My contention is that there are no known commodity resources in space that could be sold profitably on Earth.

The key to a successful business is to obtain feedstocks for cheap and to sell products at a tidy profit. The problem with space mining is that the feedstocks are generally much more expensive than on Earth, and there is an extremely limited market for products, except on Earth. More broadly, for every industrially valuable ore, there is already a competitive and adequate, if not spectacular, supply chain here on Earth.

If and when cities are built on the Moon or Mars, then local sourcing of raw materials makes sense in that context. But until then, the money, the financial resources, are here on Earth. So to make a killing in space, some sort of commodity needs to be obtained, transported to Earth, and sold, all for less money than conventional supply chains.

The challenge is that raw commodity margins on Earth are already super slim. The problem is that there are very few natural monopolies in mineral supply, so mining companies have to compete for market share, lowering prices.

More broadly, it is instructive to consider the value chain as raw materials are gradually processed into high value commercial goods, such as cell phones. Primary production obtains the ores needed to produce chemically pure elemental feedstocks, which are usually packaged in some standard, fungible way. Secondary production processes those feedstocks into individual components, such as the machining of an aluminium cell phone chassis from a raw billet. Finally, the various components are assembled, packaged, and sold. In something like a cell phone, value accrues at every step along this process, representing the revenue stream for each specialized supplier. As the designer and marketer, Apple pockets something like 30% of the sticker price of each phone sold, while the aluminium smelter takes home much less than 1%. A billet of aluminum is much closer in value to raw bauxite than a finished phone.

Similarly for minerals from space. The value per kg is of crucial importance for products where shipping costs are important, and the value per kg of nearly every commodity good is next to nothing.

But just how important are shipping costs? On Earth, bulk cargo costs are something like $0.10/kg to move raw materials or shipping containers almost anywhere with infrastructure. Launch costs are more like $2000/kg to LEO, and $10,000/kg from LEO back to Earth. Currently there is no commercially available service to ship stuff to and from the Moon, but without a diverse marketplace of launch providers, there’s no reason to expect that the de facto monopoly or duopoly of SpaceX and Blue Origin would sell it for less than $100,000/kg, literally a million times more expensive than shipping anywhere on Earth. Before we hate SpaceX for price gouging, it’s not certain that shipping for less than this amount is even possible, but one could relax this assumption by several orders of magnitude and still arrive at the same answer.

For nearly all commodities, shipping costs are a smallish fraction of the overall costs of purchase. More generally, of all the energy and labor embodied in a finished product, most of it is spent in refining, processing, design, and assembly, rather than transport. There are a handful of exceptions where shipping costs dominate the sticker price, usually in industries where transport is itself the product, and the cargo is extremely time sensitive. Shipping perishable food, flowers, and people are a good example.

Given that the Moon is not likely to (initially) be a source of perishable commodities nor enormous numbers of time-poor humans, it is safe to assume that whatever is produced there has to be so valuable on a per kilogram basis that buyers on Earth can absorb the shipping cost. The question then becomes, what commodities cost in the ballpark of $100,000/kg?

As an aside, one obvious way to sidestep the mass transportation requirement is to choose a product with no mass, such as electromagnetic radiation. And indeed, the most vibrant commercial space product is communications, which are beamed using microwaves. Raw microwaves can be used to transmit electrical power, but in a former post I demonstrated that space based solar power can’t compete with the rapid evolution of ground based solar power. Not even a little bit!

There are actually plenty of things which cost $100,000/kg or more in the high tech industries, such as advanced computer chips. The reason computer chips are so expensive (relative to mass) is that they’re extremely hard to make even at the Intel factory, which is stuffed with super smart people. In terms of the value chain, computer chips are at the complete opposite end to raw bulk commodities. Both items are sub ideal for obtaining in space, though for different reasons. Raw commodities have too little intrinsic value to justify the transport costs from space, or even usually from another continent. And high technology products are too expensive to make in any but ideal circumstances here on Earth.

There is a middle ground. The German economy, in particular, is powerfully driven by thousands of small specialty companies that make relatively small numbers of custom machines and tools. Individually, the machines are much more valuable than raw materials, and much less difficult to make than computer chips. But their true value derives from the network effect of having thousands of companies feeding off each other and, fundamentally, building the infrastructure of industrial automation for the rest of the world. There are a number of companies, such as Made In Space, which are actively pursuing bespoke in-space manufacture of specialty items, and there is every indication that their schemes are economically viable. But while they represent a golden ticket for one small engineering company, they lack a path to generalized space industry and the trillion dollar revenue that implies, at least without enormous advances in robotics.

So we’re left with a question about what commodities cost $100,000/kg, or $100/g, and could be found in space. In a previous post, we dispatched the idea of selling lunar water, which in any case is basically free on Earth. Comsats are routinely launched to space at vast expense, but fall in the category of advanced technology which is prohibitively difficult to manufacture in space. Launch may be expensive but it’s cheaper than launching the whole factory!

Let’s consider a representative list of the most expensive materials in the world. In descending order, they are:

Antimatter, currently $62.5t/g.
Californium, $25m/g.
Diamond, $55k/g.
Tritium, $30k/g.
Taaffite, $20k/g.
Helium 3, $15k/g.
Painite, $6k/g.
Plutonium, $4k/g.
LSD, $3k/g.
Cocaine, $236/g.
Heroin, $130/g.
Rhino horn, $110/g.
Crystal meth, $100/g.
Platinum, $60/g.
Rhodium, $58/g.
Gold, $56/g.
Saffron, $11/g.

The previous ballpark estimate for transport costs was $100,000/kg, or $100/g. Since I want to be inclusive, I’ll include everything down to saffron in the list above, whose cost is roughly equal to the current LEO-surface transport cost.

Despite their high value density, none of these make good candidates for commercial extraction from the Moon or asteroids, for a few different reasons.

by Casey Handmer, Blog |  Read more:
Image: uncredited
[ed. Californium? See also Einsteinium.]