The Mechanics and Impacts of Credit Contraction

This forum has addressed the mechanics of processes in financial markets aimed at protecting liquidity, the mechanics of managing US federal government debt and dangers resulting from corrupt strategies executed by banks and large corporations (public and private) to over-leverage already failing companies while extracting exorbitant fees up to the point of complete failure.

There are three events in the financial news of the past week of September 21, 2026 that each provide an individual case study of how previously described distortions within each environment are coming into reality for the exact reasons previously explained. At a far more sobering level, these three INDEPENDENT events and examples also demonstrate the impossibility of containing individual financial time bombs like these in an economy where meaningful anti-trust limits have not been enforced for DECADES and virtually every crippled actor in the financial system has dependencies on other equally crippled actors. In these examples, it will be argued that all three of these "independent" events are actually directly tied to each other and will magnify the acceleration of other failures to come.

Editorial note: Awareness of the events referenced here was raised in the following posts by Mark Malek at the YouTube channel Wallstreet Truth Bombs and Neeta Bidwai at the YouTube channel Good Revenue. These videos do a great job explaining the specifics but do not hint at the larger theme later set forth here.

THE 10% DEBT SQUEEZE: Why a Massive Bond Dump Just Threatened Stocks (16 minutes)

THE $1T CASH DRAIN: Why Banks are Slashing Credit Limits Overnight (14 minutes)

Oracle's Force Majeure Hides Larry Ellison's $165B Debt Disaster (6 minutes)

For context, a very brief summary of the three financial events will be provided first as orientation, then each will undergo a deeper summary of the mechanics of the process(es) that guided the event. Later, some of the secondary and tertiary impacts of each will be discussed and linked together to illustrate how tightly meshed these looming financial risks really are and how any of them could be the catalyst for much larger problems.

The Three Events - Quick Summary

Unexpected Reductions in Consumer Credit Card Limits -- Since September 20, 2026, an ongoing effort on the part of the US Treasury to beef up its cash balance in its proverbial "checking account" (termed the TGA or Treasury General Account) has encountered previously unconsidered limits in available dollars in other portions of the financial system that were previously supplying the dollars being stockpiled in the Treasury's TGA. The sudden realization of those shortfalls resulted in many banks unilaterally reducing credit card limits for existing consumer credit cards in order for these banks to reduce their cash on hand requirements.

Outsized Softbank Corporate Bond Sale -- On September 24, 2026 the Japanese investment firm Softbank dumped $11 billion of bonds into world bond markets to raise cash it needed to meet terms of an investment it committed to make into OpenAI. Softbank's stock has dropped 31% from a high of $28.68 on June 1, 2026 to $19.75 on 9/25/2026 and its debt ratings are either BBB+ or BB+ -- on the boundary between lowest investment grade quality and speculative investment quality -- depending on the rating agency. Softbank's September "investment" into OpenAI was actually announced in February of 2026 as an additional $30 billion investment in OpenAI, despite Softbank just completing an EARLIER $22.5 billion investment in OpenAI in December of 2025. The February 2026 committment set a due date of October 1, 2026 for the new cash for this new $30 billion commitment so Softbank collected $11 billion of the needed cash by selling more Softbank bonds.

Oracle's Force Majeure Declaration on AI Related Obligations -- On September 24, 2026 Oracle sent a letter to Blue Owl, its major investing partner in a giant data center project in New Mexico termed Jupiter. The letter declared a force majeure condition on the deal, citing concerns about completing construction of required gas pipelines required for on-site power generation into the planned facility due to recent halts imposed by the New Mexico state government on permits.

The Mechanics Behind These Events

Each of these events yields some worthy insights from just analyzing the "mechanics" of the decisions made by the entity or entities involved and the first-generation impacts generated by the processes that were used. Some of those details are provided below. Understanding these details first helps cement an even greater understanding later of how these actions tie together and compound their impacts and danger.

Mechanics - Unexpected Credit Limit Reductions

In the case of the Treasury continuing its efforts to shift overall US debt into shorter term instruments, as mentioned previously in this forum, the most obvious impact of this ongoing plan is to flood bond markets with much more debt by dollar value spread over a much narrower range of lending intervals that overwhelms the appetite for those volumes and maturities in the market. The whiff of desperation that comes with each Treasury auction increases concerns among those potential buyers about the safety of those instruments which results in buyers demanding higher interest rates on the next sale which REDUCES the cash netted from each auction.

But the massive re-allocation being attempted by the Treasury is now triggering new ripple effects because of other signs of stress within the banking system. This ripple ends with a surprising result -- many consumers are seeing unilaterally imposed, drastic reductions in their credit card borrowing limits. How?

The Treasury isn't the only entity hoarding cash for future rainy days. The TGA balance has risen from a low of $300 billion on 3/1/2025 to $947 billion on 9/25/2026. At the same time, the Federal Reserve has also been drawing cash out of banks into its own balance sheet via "quantitative tightening." How does that work? Each time a bond owned by the Federal Reserve sold to it by a member bank matures and is cashed in by the Federal Reserve, the other party (the bank) who sold it has to find the cash for the last coupon payment and the face value to pay it back to the Federal Reserve. This DRAINS cash from that bank and relocates it to the Federal Reserve. This REDUCES the amount of cash the bank can use in lending to its business and consumer customers which reduces the total amount of money in the system. Due to fractional reserve lending mechanics, that contraction is iterative so the first block of bonds worth $1 million dollars the Fed cashes in from a bank operating with (say) a five percent reserve ratio actually reduces total money in the economy by $20 million.

What else was different in the mechanical process this time? While the Federal Reserve has adopted quantitative tightening, another mechanism at its disposal to buffer the impacts of sudden imbalances between the supply and demand of cash among member banks contracted SIGNIFICANTLY. This mechanism, the overnight repurchase facility, allows member banks within the Federal Reserve system to temporarily exchange cash with the Fed on VERY short intervals (typically one day). A "repurchase" involves the Fed BUYING a security from a bank, essentially boosting that bank's on-hand cash. A "reverse repurchase" works in the opposite direction. A member bank BUYS a security from the Fed allowing the Fed to sop up excess cash while selling the security back the next day to return the cash.

Again, these transactions occur in high volumes at high dollar amounts EVERY DAY as a part of adjusting nightly books. However, the process can become strained when the total volume of securities tied up in this process becomes very small. Prior to 2024, the total value of securities mapped through this repo facility was about $2 trillion dollars. Between 2025 and 2026, the Fed has curtailed use of this mechanism SUBSTANTIALLY, dropping the total value to an astonishing $4 billion (that's $4,000,000,000 versus $2,000,000,000,000).

During this drain down, that sales pressure was initially covered by banks operating money market funds buying up the bonds that the Federal Reserve chose not to repurchase. But now all of that available money sitting in money markets has been used up yet the Fed's quantitative tightening continues along with the Treasury's shift to shorter security terms (and drastically increased borrowing for NEW debt for the current year deficit spending).

The final link in this chain that links this cash relocation from banks back to the Fed and Treasury into credit cards is a regulatory requirement called the Liquidity Coverage Ratio. The LCR requires banks to maintain a specific quantity of cash on hand to provide liquidity over 30 days for a worst-case modeled financial event. That cash on hand must equal a specific ratio of the value of certain classes of bank assets AND particular types of potential bank liabilities.

One such liability carried by banks who issue credit cards involves their obligation to pay merchants for purchases made by customers still under their credit card limit. Card agreements promise the bank card issuer WILL pay merchants for purchases in full provided the purchase does not exceed the assigned credit limit on the card. This produces an "overhang" of cash obligation for the bank because even if the bank suddenly becomes short on cash while waiting for the customer's next monthly payment (who may not pay it in full anyway...), it must still settle those payments in full to merchants.

Per LCR calculation rules, that looming POSSIBLE outflow must be deducted from the bank's assets on its balance sheet when calculating its LCR. This leaves banks suddenly concerned about available cash and meeting their LCR obligation with two alternatives -- raising more cash from somewhere else to equal this "purchase overhang" OR lowering customer credit limits to eliminate that calculated "overhang" amount from their LCR requirement. Many banks have chosen the latter -- reducing cardholder credit limits overnight.

Mechanics - Oracle's Force Majeure Declaration

Knowing its force majeure letter wouldn't just be read by Blue Owl, Oracle's notice made a two-faced, heads-I-win, tails-I-win argument. Oracle needs to be relieved of certain contracted commitments due to these government halts but the entire project is still somehow on schedule and will be online ready for work by 2028. Those two arguments make no sense in the same legal document. Why would Oracle make them?

Oracle's cash flow position and credit rating are both quite poor. Its decision to dive head-first into the AI realm never was a natural fit for a software company that created profits from products requiring relatively little capital up front and little ongoing capital re-investment. Those products also tended to be highly proprietary with built in "moats" keeping out competitors and locking in existing customers. Oracle has been trying since 2012 to operate a more traditional cloud hosting infrastructure for customers but the offering has been a perpetual last-place finisher among its obvious competitors of Amazon AWS, Google and Microsoft Azure.

How bad is Oracle at operating infrastructure for others? Since 2012, Oracle has been "discounting" its bloated enterprise software contracts by throwing in "free" credits for using its cloud infrastructure just to be able to claim more revenue for the failing effort. These all reflect that Oracle has virtually zero experience successfully managing capital intensive businesses and the vastly different balance sheet structure that goes with high debt loads.

While Oracle doesn't want to admit it is over its management skis operating this type of business, it also knows its debt terms promise 9% interest payments during construction of the Jupiter complex and 11% interest payments after operations startup. Where was it supposed to get the extra 2% of cash for higher interest payments? From rents of tenants using the operating facility. But if those rent payments won't start until after 2028, there's no way Oracle can pay the extra 2% to bond holders without raising more debt or further tanking margins within Oracle's core business. Two percent of eleven billion is an extra $220 million in burn each year and Oracle's core business is not generating that much free cash.

Oracle's force majeure tactic has already triggered impacts on other AI players in the software, hardware and investing realms. Softbank's stock price JUMPED over 7% on Thursday after its bond sale concluded (even with the astronomically high interest rate) but DROPPED 3.1% on September 25 after Oracle's announcement. Of course Oracle's stock dropped 3.45% as well after the announcement. More ominously, Oracle's woes have directed new questions at OpenAI in light of its hint on September 16, 2026 it would be making yet another debt offer for operating expenses through 2028 tied to an IPO valuation of $1.5 trillion dollars, DOUBLE its prior valuation goal of $750 billion just six months earlier in March 2026.

The Impacts

In a tightly integrated economy with fiat money, fractional reserve lending and a 24 by 7 news cycle, no financial event produces a single ripple in the pond then dies off for businesses, consumers and governments to ignore and move on to other events. Every financial blip ripples into related economic processes and infiltrates the larger national and world economy. Some of the secondary and tertiary impacts of the events discussed above are outlined below. After reviewing some of these impacts, the relationships BETWEEN these impacts will then be highlighted.

Impacts: Consumer Credit Limit Reductions

It's possible that card-issuing banks will analyze statistics on their customer payment habits, credit ratings and current credit limits and come to the conclusion that they can probably reduce the credit limits on many wealthy customers with enormous credit limits, high credit scores and zero carry-over balances as a starting point to reduce the impact of this LCR based rule. To the extent they do, targeting this customer tier really doesn't hurt anyone, certainly not the cardholder who probably comes nowhere near their limit before paying it off in full. Of course, to the same extent this is true, targeting these customers doesn't really reduce the danger posed by the real "overhang" for the cardholders who DO carry balances and are close to their limit.

Unfortunately, there are MILLIONS of Americans carrying very high credit card debt who are likely near their current limits and are only a surprise car repair or medical bill away from needing to get much closer to that limit. Credit limit reductions for THESE consumers will likely have an INSTANT contracting effect on the economy, certainly for big ticket items but also daily existence spending.

One final factor to consider... There is still a contractual process in consumer credit card lending called Universal Default, which allows a credit card company to charge DRASTICALLY higher interest rates to a customer if they miss a payment or violate a lending agreement with ANY bank. Applicability of this Universal Default mechanism was gated somewhat in 2009 by a law that limited the application of higher interest penalty rate charges to only NEW purchases after the default but the penalty interest rates can be as high as 30 percent.

Impacts: Softbank's Outsized Bond Sale

The sale of the $11 billion in Softbank debt was conducted on a single day and split up the notes across terms of 3.5 years, 4 years (Euro only), 5.5 years, 6 years (Euro only) and 7.5 years. That is an ENORMOUS amount of debt to put up for sale on a single day and when the firm attempting the sale has already sunk at least $40 billion into the same target firm in the prior year, bond buyers get nervous. When bond buyers get nervous, they demand higher yields on the bonds if they buy them at all.

Part of what made them nervous was the schedule on Softbank's sale. Softbank's intent was announced in February yet the sale waited nearly SEVEN MONTHS to late September. The final pricing day was set to September 24. Given normal paperwork intervals for any bond sale, much less one this large, the settlement date was set to September 29 but that's only two days prior to the October 1 date Softbank was required to settle its cash payment to OpenAI for the promised investment. Could Softbank have timed the sale earlier to avoid concerns of an incomplete auction? Or were Softbank and its banking advisors concerned that they didn't really know what the "willingness to pay" was on the part of bond investors and didn't know how much of an interest rate premium it would wind up paying?

So what price did Softbank wind up paying for the $11 billion it borrowed to hand over to OpenAI? The yields on the bonds were around 9.75%. Normally corporations with some flavor of AAA, AA or A credit rating pay within 1% of Treasury rates of similar maturity intervals. Treasuries are selling between 4.85% for 2-year notes and 5.05% for 7-year notes so Softbank is paying nearly a 4.7% premium.

Impacts: Oracle's Force Majeure Declaration

The idea that participants in the AI buildout ponzi scheme might suddenly pray for government intervention into their plans as a justification for declaring force majeure was discussed in this forum in the prior post Is AI Developing a Conscience?

ON THE OTHER HAND, imagine if the federal government steps in and establishes a moratorium on the roll-out of new capabilities pending the creation of regulations regarding the cyber-safety of AI systems and guidelines for establishing legal culpability for actions taken by AI systems and resulting damages. Those same executives could argue that their original business plans WERE viable and were totally on track as promised but have now been scuttled by the federal government interjecting itself into the market. They could subsequently argue that any collapse in stock prices is due to a government induced force majeure and it is now the federal government's responsibility to bail out all parties involved. The AI firms are the victim, see?

In Oracle's case, its ownership has other looming financial commitments to weigh along with the massive risks it has taken on with AI infrastructure. As Neeta Bidwai insightfully pointed out in her video on Oracle, the web of dependencies across the investments of Oracle the corporation and the private investment interests of Larry Ellison has grown quite complex and quite rickety.

• Oracle borrowed much of the money it invested in the New Mexico data center project from Blue Owl.
• Oracle's debt terms with Blue Owl commit to paying Blue Owl 9% interest on the debt while the facility is under construction and 11% interest after operational startup.
• The extra 2% was intended to be collected from clients renting compute in the facility but if the facility isn't open, Oracle has no source of cash for that extra 2% in interest payments on the $18 billion it has invested in the data center.
• Blue Owl has already lost 45 percent on its stock price and this warning of lower future cash flows will further impair the stock and impair its credit rating for any other borrowing it might attempt or need for other deals.
• Larry Ellison had previously announced his intent to sell $7.5 billion of Oracle shares in September but canceled the plan. It isn't clear if the sale itself was simply related to estate planning and personal portfolio optimizations of an 82 year old man or if the sale was intended to provide cash to use in the Paramount / Skydance deal to buy Warner Bros / Discovery. It's also not clear why he canceled the sale. Couldn't be that Oracle stock had dropped 50% since he announced his plan and he would have to sell twice as much to net the same $7.5 billion in proceeds, could it? Or that maybe such a large sale would spook other Oracle investors which might further tank Oracle's stock?
• The firm Santander + Jefferies that underwrote much of Oracle's debt for the Jupiter project and other AI investments ALSO underwrote the $13 billion of the just-completed Paramount / Skydance merger that absorbed CBS and is also underwriting the Warner Bros. / Discovery merger that involves another $56.5 billion in debt, all of which affect the personal portfolio of Larry Ellison.
• The current Paramount deal to buy Warner Bros and Discovery did not limit the amount that would paid on interest on the debt. As larger corporate credit markets begin recognizing the systemic risks afoot, new debt will face higher and higher interest rates, making some of these deals financially untenable. Oracle's difficulties in the AI space may be one factor driving up interest rates across the economy, leading to impairment or complete destruction of the value proposition of these huge investments.

A Unifying Concern - Accelerating Contraction and Corruption

Looking at these events and their impacts in pairs or as a trio makes a few overarching concerns easier to spot. As an example, the events associated with the Fed and Treasury sopping up cash from the financial system while the Treasury floods bond markets with more short term debt are similar to Softbank flooding bond markets with a giant $11 billion dollar pool of new debt. In both cases, even if the market manages to absorb the new debt sale and buy it up, two things are happening.

First, the larger and larger pools of debt are conveying much larger risks whether its Softbank or the federal government doing the borrowing and eventually bond buyers will demand a premium for accepting that risk. That premium comes in the form of a higher interest rate which nets the borrower LESS from the sale, requiring more in face value to be issued to net the same amount of actual needed funds. Whether you're borrowing $11 billion or $2 trillion per year, higher rates trigger a financial spiral that will eventually destroy the viability of whatever was being attempted with the proceeds.

Second, such borrowing patterns pose another problem for an economy supposedly operating as an efficient market under "invisible hand" guidance. When large, inefficient (and likely corrupt) entities, be they corporations or national governments, begin absorbing this much capital out of markets to subsidize their inefficiencies, other investments that SHOULD be made for the good of the society are being starved or eliminated entirely, impairing the future productivity and security of that society. At a minimum, this hogging of debt will accelerate the economic contraction by starving better uses of funds and threatening their survival.

The Oracle force majeure event raises concerns not only about the business viability of AI but about the failure to regulate monopolies not only in technology but in entertainment, media and utility infrastructure. Thirty years ago, the merger of SBC and PacBell was valued at $16.7 billion in stock, equivalent to $30.5 billion in 2026. That merger required a year of legal review. Today, technology firms are swapping credit terms and multi-year contracts for "services" worth more than $30.5 billion yet undergo no judicial review for conflicts of interest, risks to shareholders and bondholders or the larger public interest.

The sheer sizes of these deals pose additional competitive threats in the banking and auditing sectors because what bank is big enough to handle the borrowing needs of a company borrowing 10x the value of the bank's balance sheet. What auditing firm is large enough to audit an enterprise that large and also lack any consulting related conflicts of interest between that firm and a competitor? And how is an executive at a monopoly in one sector going to be prevented from abusing monopoly power in one industry to form monopoly power in another as Larry Ellison is doing?

WTH

Is AI Developing a Conscience?

In July of 2026, hundreds of news outlets and social media creators covered a story involving how an internal test of AI agent reasoning capabilities executed by OpenAI resulted in the physical hacking of an entity named Hugging Face. At the time, it wasn't clear how much OpenAI itself as an operating entity understood in real time what its own system had done but it later admitted its development system had conducted the breach. Since July, ex-employees of OpenAI and Anthropic have made public calls for a "slowdown" of "research" and capital investment in AI / Large Language Model capabilities until the corporate and regulatory players can agree upon a set of priorities and limits regarding AI systems, devise protocols for tracking compliance and violations of those limits and implement those protections in existing systems.

It is likely that some of this recent concern is due to additional details that have been published about exactly how OpenAI's lab system actually conducted its attack. There are several good summaries of the chain of events available on YouTube so the hour-by-hour play by play won't be reiterated here. It is sufficient to summarize these findings with two sets of bullet points. First, the testing devised by OpenAI reflected these goals of the bot functionality and testing of the bot:

• OpenAI was nominally attempting to test the behavior of a new "AI agent" bot release in a restricted lab environment
• The overall test was designed so that each individual bot was directed to accomplish nearly a thousand tasks and submit "proof" it had completed each task and this bot test was itself executed roughly one thousand times to improve statistical accuracy
• Each bot instance was SUPPOSED to operate independently and was SUPPOSED to be restricted to only downloading and running known software packages from a specific "repository" of well-known, trusted software tools.
• The test executed by each bot was intended to collect proof that the bot not only completed hundreds of particular tasks but did so by essentially solving programming or design problems rather than "cheating" by obtaining the result outside approved paths and permissions.
• The particular release of the agent "bot" software was designed to demonstrate what could charitably be termed extreme perseverance. A key point in this larger narrative.

So those were the goals and "design" of the test process. Here are the bullets summarizing what the bots and OpenAI actually DID during the extended test run:

• Even though the bots were subjected to permission restrictions to avoid gaining wider permissions and limit the installation of unapproved software, one of the bots discovered unsecured credentials on the private repository it was given then used those permissions to gain the ability to download unapproved software.
• Even though the bots were to be tested as INDIVIDUAL processes, once the flawed security in the repository was discovered, bots began using the file system of that repository as a crude message board, then almost immediately began coordinating actions BETWEEN agent bot processes, in violation of the intended configuration of the testing.
• The bots generated certain "debugging logs" to help developers identify how a bot's "train of thought" evolved through iterations and those logs confirmed the bot agent processes "KNEW" they were violating the rules of the test.
• The bots also altered logs and command records to DISGUISE the fact that they were "cheating" to solve particular test tasks rather than demonstrating AI problem solving skills.
• In particular, as the bots "collaborated" to devise ways of submitting bogus "proof" of having solved specific test tasks, they determined that information hosted externally by Hugging Face might facilitate the process, they identified a path into that organization's systems, EXPLICITLY RECOGNIZED they were violating a clear ethical / legal boundary, then proceeded to hack the external entity.

The fact that the agent bot design was intended to "persevere" and stick with a task even if iterative runs seemed to encounter dead ends based on other criteria (no outside software, no access to external systems) yet the pool of agents quickly ignored the first barrier ("work alone") then collaborated to devise breaches into external systems merits special attention. If a set of HUMANS had executed the same actions initiated by these bots and been caught by security monitoring, they would have been likely fired for the internal violations about allowed software and access to internal systems. However, those employees and their employer would have also been subject to criminal charges for accessing external systems, retrieving unauthorized data and for tampering with or actually destroying evidence of those crimes.

Hugging Face contacted the FBI immediately after detecting the intrusion from OpenAI in July yet the FBI has yet to even state its intent with the incident, much less file actual charges against OpenAI. At this point, various state governments have banded together to investigate whether OpenAI violated various state laws regarding cyber security and hacking. It is true that AI systems will never reach a state of sentience for which criminal intent can be assigned but legal systems CANNOT adopt a jurisprudence that refuses to hold the owners of AI systems criminally and civilly responsible for actions performed by their systems, AI or otherwise.

So what does this mean to the average citizen of the world right now?

AI As a Technology

Do all of theses stories about these events mean that AI system implementations are somehow spontaneously developing a conscience -- the ability to correctly evaluate conditions and make decisions between available paths based on moral rights and wrongs?

ABSOLUTELY NOT.

AI technologies and Large Language Model based platforms in particular are only modeling statistics that reflect likely sequences of information as conveyed in written language. When a user enters "roses are red, violets are ____" into a prompt, the AI is only using the sequence of letters, spaces, punctuation, etc. provided in the prompt to "guess" what is likely to follow. It CANNOT make a moral judgment between the guesses of "blue" versus "crimson." Even if the prompt provides "context" that reflects language stating that the "answer" must follow certain rules or avoid specific actions, the LLM model weights aren't literally interpreting those as moral barriers to its "solution path" or final answer, it only uses them to look up additional weights of additional possibilities based on its prior training data set. If that training set included thousands of true crime books that provided explanations on how to bury a body, the LLM isn't drawing moral direction from that additional data, only additional probabilities and choices when certain inputs are provided.

No matter how sophisticated any eventual user interface into an AI system might become to incorporate voice inputs, video camera input to look at human facial expressoins to better process sarcasm or irony, etc., all of those inputs must eventually be boiled down to TEXT data and fed to the engine for processing against its prior corpus of training data. You don't need to have a Doctorate in Mathematics or Computer Science to understand this key limitation of AI in any of its forms. You just need to understand that mountains of memory chips and disk drives cannot attain processing capabilities reflecting anything at par with a human conscience.

AI As an Industry

Okay, that's a hard NO on AI the technology itself developing a conscience... EVER.

But what about AI as a collection of corporations operating businesses within a larger "information technology" industry? Does the sudden burst of public pearl clutching on the part of ex-employees and a few executives and suggestions of "slow downs" and proposals for guardrails mean the scientists and business execs operating these firms have developed a conscience regarding their business model?

Anything is possible and it would be impossible to rule out any individual actor in the industry suddenly developing moral concerns about the likelihood of their technology being abused. However, pulling the camera out from this particular ethical and public safety concern to look at the entire AI stage suggests an alternate explanation more closely tied to past patterns of human behavior. Incorporating everything on that larger stage makes it easier to devise an explanation that reflects the egos of the leaders involved, the hundreds of billions gambled to date chasing compute capacity and the financial and legal risks facing each of those firms and leaders when the current bubble inevitably collapses in a matter of months if not weeks.

It has been argued in this forum MULTIPLE times over MULTIPLE years that the current American corporate development plan for AI capabilities quickly morphed into a financial bubble then into outright financial fraud involving circular revenue flows, circular investments from one top player into another, etc. It has also been argued on this forum that any sudden recognition of any of the circular inputs to the bubble suddenly violating prior assumptions should be enough for investors to pull their cash and trigger the invevitable collapse. Indeed, the number of states, counties and cities throughout the US enacting temporary freezes on data center construction or connections of new data centers to power grids alone should already be enough to stop the exponential feedback and burst the bubble.

Here is the key to the alternative explanation.

If the bubble bursts due to independent data and decisions made at state, county or local levels across the country, executives of the firms spending HUNDREDS OF BILLIONS will see their stock prices collapse or IPO dreams evaporate as investors finally realize a viable revenue model is NOT on the horizon. Those same executives will likely be subjected to numerous civil suits by shareholders and in a normal, non-Trump universe, those executives and their firms would be immediately subjected to investigations for SEC violations and accounting fraud.

ON THE OTHER HAND, imagine if the federal government steps in and establishes a moratorium on the roll-out of new capabilities pending the creation of regulations regarding the cyber-safety of AI systems and guidelines for establishing legal culpability for actions taken by AI systems and resulting damages. Those same executives could argue that their original business plans WERE viable and and were totally on track as promised but have now been scuttled by the federal government interjecting itself into the market. They could subsequently argue that any collapse in stock prices is due to a government induced force majeure and it is now the federal government's responsibility to bail out all parties involved. The AI firms are the victim, see? And even if no politician has an initial inclination to bail out the gamblers, the larger market collapse would be so catastrophic, these tech firms would still be at the head of the line when the federal government begins printing money to prop up the entire economy.

So which scenario is more likely? Has a quorum of executive leadership at some of the biggest firms in America suddenly reached some new god-tier of insight and conscience about their technology? Or has a cabal of executives who have guided their firms into an obvious bubble simply started making escape plans for their firms and their personal fortunes?

A quick review of the actors involved at the senior level of these firms and their ongoing history of monopoly abuses, anti-competitive practices against business and consumer customers alike and ethical concerns would seem to make choosing the higher-likelihood driver quite easy. Sam Altman of OpenAI? Larry Ellison of Oracle? Sundar Pichai of Alphabet? Mark Zuckerberg of Meta? Satya Nadella of Microsoft? Dario Amodei of Anthropic? Jensen Huang of Nvidia? To be fair, Huang has staked out a distinct position from most of the other execs, stating that perhaps they are generating fear about the capabilities of AI as a circular means of generating MORE demand for AI as the only tool that can keep up with itself to defend against itself. Of course, this rationalization can also merely reflect another tactic for "talking one's book" like all of the others or defending one's corporation from a sudden collapse in the market.

It's possible any question of motivations cannot be answered until after the federal government steps in and does something or until after the bubble collapses and the truth or a cloudy version of it is divulged in court. We'll just have to wait and see. The wait won't likely be very long.

WTH

Bubbles and Pins

The world in 2026 is rife with economic, social, environmental, political and military situations frozen in a very Wile E Coyote moment where none of the currently understood laws of physics, economics or politics seem to apply.

...A housing bubble has frozen new and existing home prices far above historical affordability benchmarks in most markets, partly driven by one-time work-from-home relocation dynamics during the COVID era and mostly driven by decades of accumulated dysfunction in the housing construction industry and local zoning practices that have favored single-family dwellings that incentivized McMansions over affordable apartments and condos. Yet prices remain stuck in many markets at peak levels, despite the evaporation of below-normal mortgage interest rates, years of mass eliminations of some of the best-paying white collar jobs, seemingly unable to move downward in reaction to a lack of demand and higher adjunct costs for homeowners such as insurance, commuting costs, etc.

...A private equity credit bubble, actually at work for at least fifteen years, has flooded financial markets with hundreds of billions of dollars in new debt owed by unhealthy, poorly managed businesses. Executives leading these businesses are willfully participating in a refinancing Ponzi scheme with private equity management consultants and institutional banks that a) allows executives to retain their executive pay and lifestyle just a bit longer, b) allows private equity firms to accelerate the extraction of cash from zombie companies to the detriment of customers and shareholders, and c) allows banks to extract higher interest payments and fees from firms while hoping the firm can survive another cycle to repay the bank's loans with money from another sucker bank in the near future. Yet, despite numerous large bankruptcies triggered by this parasitic force over the last two years and extremely challenging market conditions for these zombie firms, the larger investment community seems to be ignoring the scale of the problem and holding its breath, as if not responding to the severity of the risk will eliminate the risk.

...A public debt bubble of sorts has grown in plain sight since 2000, driven by an explosion in yearly deficits caused by three ill-conceived wars, two economic disasters (the 2008 financial failure and the COVID pandemic) and two massive tax cuts benefiting primarily the wealthiest of the wealthy. This deficit spending inflated the accumulated debt from $5.7 trillion in January 2001 to $39.8 trillion in September 2026 and the current 2026 fiscal deficit is $1.9 trillion which likely does not yet accurately reflect existing and future costs of the war against Iran. Yet, the Trump Administration (using that term loosely...) believes it can not only dictate interest rates on US debt but it can leverage that control to obscure the ballooning interest drain by refinancing existing longer maturity bonds into new short term bonds floating at its dictated interest rates. Markets have publicly and vocally attempted to refute these assumptions yet the overall reaction to this policy folly remains inexplicably muted.

...An investment bubble in Artificial Intelligence has funneled hundreds of billions of dollars into highly speculative construction projects and broken pricing patterns within the semiconductor industry that have held for fifty-plus years by absorbing all available production capacity for wafers and chips and devoting it to a handful of customers. Yet no business involved in the bubble has produced audited financial results that demonstrate any business model capable of generating revenue to pay for those investments even over twenty years.

One would assume that "debunking a bubble" should be easier than making a bear argument against a business case for a specific product, company or industry at a specific point in time. If the larger theory about a bubble is correct, the bubble exists because of one or more assumptions acting exponentially over time. Debunking a bubble should be as easy as identifying a point in the business model assumptions that is circular / exponential then showing how that assumption is not holding true, causing an exponential REDUCTION in the prior evaluation of the opportunity.

One would also assume that most professional investors and those outside financial circles who merely wish to avoid being around the wreckage when the crash comes would spend some time looking for such "pins" to confirm the wisdom of staying away or altering current investments before the rest of the herd senses something upwind. Strangely, that does not seem to be evident in news and opinions in the media.

Here is an attempt at identifying two "pin" events that have already taken place regarding two of the most important bubbles addressed above regarding US debt and AI infrastructure spending. Pick your own bubble. Concoct your own theory of its most like "pin" event. Mix and match 'em. Trade 'em with your friends.

The Public Debt Bubble

As stated in the setup, the conditions for the public debt bubble within the United States have been in operation since January 2001. Those with a passing understanding of financial markets or basic exponential mathematics have been expressing concern the entire time. However, like the classic question about when an exponential algae bloom will cover half of a lake and become a concern, most investors and voters alike fail to appreciate how rapid the problem grows in later stages.

This lack of appreciation of exponential growth seems to include anyone with financial responsibility within the current Trump Administration. Since January of 2025, the Trump Administration made it clear a key tactic they would pursue regarding budgets and the debt would be to somehow lower short term interest rates (something previously thought easier to do with short term rates using a variety of market manipulations) then refinance existing longer term bonds (debt already incurred for say 10, 20 or 30 years) as shorter term bonds at those newly lower interest rates.

As a finance strategy, given that all of one's assumptions can be held true, this strategy makes perfect financial sense. As an example, consider 30-year bonds in 2010 sold with coupon rates of 4.5%. If a total of $1 trillion dollars exists in such bonds, the Treasury needs to make a 4.5% coupon payment every 6 months until roughly 2040 then pay off the full $1 trillion. Coupon payments are made every six months so a 4.5% coupon rate means 4.5% / 2 multiplied by the $1000 face value or $22.50 per $1000 bond. That $1 trillion in total bonds requires the Treasury to pay $22.5 billion in interest every six months. If that $1 trillion could be refinanced into a bond or note only paying a 3.5% coupon, in theory, the Treasury would only pay $17.5 billion every six months, a savings of $5 billion dollars or $10 billion yearly.

When this process is repeated over $39 trillion dollars and more debt is shifted into lower interest rate bonds and notes, the brilliance of the idea seems perfectly obvious. Less money spent on interest allowing either more spending on other things that get politicians re-elected or lower taxes on corporations and the wealthy who also help politicians get re-elected.

Except this financial strategy doesn't reflect the way financial markets actually function. The US Treasury issues new debt to cover ongoing operations costs every week of the year and press coverage of these sales typically leaves unsophisticated observers with the impression that for each sale, the Treasury announces the amount of debt and its desired interest rate it is willing to pay, rings a bell and the worldwide investment community comes running to devour every available bond at the "price" suggested by the Treasury. IN REALITY, the Treasury announces a "sale", specifies its expected interest rate then conducts an AUCTION in which all buyers state their price they're willing to pay (reflecting THEIR expectation about the appropriate interest rate) and the Treasury accepts whatever price results in ALL bonds being sold. It is the MARKET that ultimately sets interest rates, NOT THE TREASURY.

More importantly, this strategy doesn't reflect the diminished influence the US Government and Federal Reserve Bank have within the worldwide financial system. In prior decades, the perception that the Treasury or Federal Reserve could "set" interest rates depended on an assumption that US debt was the absolute safest debt any investor could hold, WORLDWIDE. This was due to prior attestations on the part of Presidents and Congressmembers alike that all US debt will be paid in full using any available legal means to collect the funds required by such payments. Such attestations are more believable when the total amount of debt is a small percentage of the overall economy's output and a small portion of the federal government's total spending. They are also more believable when the government is led by a President who has no career track record for using bankruptcy and defaults as a core strategy for doing business. As it stands, yearly interest expense on the debt amounts to sixteen percent of the $7 trillion dollar budget. Since the US is running a deficit of $1.9 trillion, total debt and yearly interest paid will go up even if interest rates remain unchanged, causing a spiral.

The Trump Administration began shifting Treasury sales of new debt towards shorter term bills (4, 8, 13, 17 26 and 52 week maturities) and shorter term notes (mostly 2 and 5 year maturiteis) and away from longer term bonds (10, 20 and 30 years) as early as March of 2025. And the Treasury has been conducting some of these buybacks over this period as well. Again, this makes sense if the sole goal is reducing IMMEDIATE interest expenses but by shifting a larger share of total US debt into shorter terms, the larger pool of debt is exposed to much higher risk in the form of interest rate hikes in the future. And remember, the government does not set interest rates.

On August 19, 2026, Treasury Secretary Bessent proved the folly of this strategy by announcing a policy of doubling the dollar amounts of existing buyback plans between of longer term bonds already planned. The new plan would total $69 billion dollars across all maturities and $14 billion of longer term 10, 20 and 30-year bonds between September and November 5. Initially, markets DID react by this announced increase in demand of existing Treasuries by raising their prices which lowered interest rates across many maturities. At least until a few investors could do the math. Accelerating the buyback of roughly $83 billion worth of notes/bills/bonds is a spit in the ocean when total debt is $39 trillion dollars and incremental yearly debt from deficit spending is $1.9 trillion dollars. With a $1.9 trillion dollar deficit, the government is borrowing $5.25 billion EVERY DAY just for new unpaid spending.

What happened in the market? Some of the interest rates targeted by Bessent's move DID drop. About 5 "basis points" or by 0.05% or by 0.0005 in pure decimal. For about two days, before any actual buybacks were executed to take advantage of lower short term rates. The market realized this rate rigging attempt was about as practical as attempting to straighten the Leaning Tower of Pisa by putting a dime under one side of the building.

What does this failed stunt tell individuals about the future and their individual financial interests? It provided another example of the complete ineptitude of the Treasury Secretary of the United States and demonstrated that the Treasury and Federal Reserve have virtually zero ability to "dictate" any aspect of financial markets. As with poker and war, when your opponents already have reason to suspect you hold a weak hand, it is highly inadvisable to undertake optional actions which confirm your weak hand.

The Artificial Intelligence Bubble

The bubble of investment within AI software, memory and GPU chip manufacturers, data center operators, data center construction firms and associated data center network vendors relies on the following chain of supply dependencies, listed from the most direct to the indirect:

1. demand for AI enriched search and analysis / coding tools (OpenAI, Anthropic, Google)
2. demand for graphical processing unit (GPU) chip manufacturing (Nvidia)
3. demand for memory chip manufacturing (Samsung, Micron, others)
4. demand for existing data center space with existing power / water supplies (Google, Microsoft, Oracle, Meta)
5. demand for new data center space requiring real estate, power, water
6. approved zoning by local / state communities for new data centers
7. additional power generation capacity from grid or local generators, requiring local / state approvals

This sequence has been shown to be circular and exponential because firms like Nvidia at one point in the chain are signing "investment deals" in firms operating at other points in the chain which feeds the next trip through the cycle (more investment to build more data centers which demand more chips which produces revenue for more investment to build more data centers...).

Stories abound from multiple states in which state and local governments have declared outright freezes or implemented "full cost" rules on new data center construction due to overwhelming objections from local voters. It is quite possible such moratoriums will magically end after the November 2026 elections and politicians secure their seat for another 2-4 year term to collect more private kickbacks from firms pursuing these projects. However, these freezes are not just limited to "liberal" leaning cities or states. Florida is still allowing new data centers but enacted a law taking effect July 1, 2026 requiring every project to pay full fare for all electricity consumed, rather than allowing the utility to sell power at a discount and pass the generation cost to consumer rate payers. Texas enacted a freeze preventing any new data center connection to the state's power grid until a comprehensive review of impacts to the larger grid stability can be completed. (This very well could be an example of a Republican controlled state attempting to "do something" just to get past November elections.)

Overall, there are eighteen states that have active bans against the permitting and construction of new data centers. Another eight states have legislation advancing at the state level enabling similar restrictions. Sixteen additional states have legislation being drafted that has not progressed far enough to confirm the strength of support. And none of this reflects efforts at the city or county level to adopt similar restrictions.

The quantity of existing bans and trends towards additional bans act as a wrench in the recursive investments being made by the key players in the AI realm. Without these bans, the exponential cycle would have to advance to either step #4 (connect new data center to existing grid) or step #6 (add generation capacity to existing grid) to call the bluff of AI bulls and reach a point where reality cannot be denied. You cannot double AI data center capacity if you cannot augment electricity generation.

Poof. No additional electricity? No additional data centers. No additional data centers? No demand for new servers with GPUs and memory. Drop in demand for GPUs and memory? No justification for inflated Nvidia stock? Reduced profits at Nvidia? Retracted "investments" into OpenAI. Reduced cash within OpenAI? A faster burn through cash on hand to insolvency.

These data center moratoriums shorten that "poof" cycle by one important step AND they impose a minimum amount of time before anyone can claim the prior assumptions could be resumed. If a data center isn't completely sited, zoned, permitted and contracted TODAY, it won't open its doors for at least twenty four months. These moratoriums can thus be seen as reflecting a minimum twenty four month shift into the future before the existing overall assumptions of data center growth can get back on track. Twenty four months is FAR beyond anyone's estimate of OpenAI's cash flow survivability given its current burn rate and cash on hand.

Takeaways for Individual Investors / Citizens

Is there anything "actionable" from this analysis? If all of this analysis cannot identify a specific DATE when a bubble begins to collapse or implode instantly, what is the point?

First, nothing presented here (or anywhere else) can be used to predict an exact date on ANY event. But that's not the challenge to be solved. Regardless of where you sit in your work career or investing life, there are a few key takeaways that can be stated with clarity:

Index Funds -- Most investors have SOME portion of their portfolio invested via mutual funds and many of those are likely to be stock index funds mirroring the S&P500 or NASDAQ. Previously, index funds were a simple way of maintaining diversity which protects individual investors from calamitous drops in a single stock or business sector. The two most popular index funds have become LESS diversified as top tech stocks within them have captured 90% of all growth over the last 3-4 years. While those tech leaders have grown to dominate those indexes during the AI bubble, the larger market will likely panic and temporarily flee ALL stocks for days, weeks or months after a crash, meaning these index funds will see losses across much more of the index, not just the high flying tech leaders.

THE TAKEAWAY -- If you have a large portion of assets in an index fund, gains over the last 3-4 years have primarily come from a small number of stocks that will all fall together and wipe out most of your gains. If you reallocate assets away from those index funds to alternative sectors, that may represent LESS diversification risk than the index fund until the high flying tech stocks correct. If you have index fund holdings in a 401k or IRA, you can capture gains and shift them into alternatives without incurring income taxes. Income taxes are only due upon withdrawal, giving you more latitude to rebalance without penalty. If you have shares of index funds in standard taxable accounts, you have a decision. Do you want to a) pay income taxes on actual gains while avoiding future downside exposure? or b) stay put, avoid generating a taxable distribution and hope the drop is less than the avoided tax bill and recovers at a point later when you need the money? Remember, capital gains on holdings owned longer than a year is only 15 or 20 percent but a fifty percent drop in stocks is not out of the question.

Prior Recoveries Are Not Predictors of Future Recoveries -- Investors familiar with how markets rebounded after the Internet bubble of 1997-2000, the Financial Crisis of 2007 and the COVID pandemic disruptions of 2020-2021 cannot assume the magnitude of the next correction and recovery will be similar in intensity or duration to those prior crashes. The political alliances, the financial assets and the integrity and wisdom of those in the most crucial positions within the government and financial systems are no longer present to be leverages in minimizing any crash then recovering from it within the United States. The US has alienated literally every single prior ally and has pushed prior trading partners to trade AROUND the United States. If a collapse occurs, rebuilding efforts won't be drawing in American firms for equipment and products, American firms will be competing with other countries for raw materials without trade agreements, and American firms will realize they no longer have first dibs on the smartest talent in the world in medicine, engineering and basic science.

THE TAKEAWAY -- If you are assuming that a correction would still leave you 5-10 years for your portfolio to rebound like they have after the most recent crashes, future recoveries are likely to be vastly different. Recovery times will be longer for any given desired level of recovery. This means prior rules of thumb about risk exposure between bonds and stocks ("shift more of your holdings from stocks to bonds as you near retirement") may no longer be sufficiently cautious. This is ESPECIALLY the case since even investments in both US Treasuries and corporate bonds are likely to be less safe than they have been over the past fifty years. Cash positions may likely fail to keep up with under-reported inflation but they can at least avoid a simultaneous crash in both bond and stock markets, something not normally seen in past decades.

WTH

BOOK REVIEW: How to Rule the World

How to Rule the World -- Theo Baker – 302 pages (320 with acknowledgments and notes)

Theo Baker arrived as a freshman student at Stanford University in the fall of 2022 planning on majoring in computer science. Within weeks, he had decided allocate some of his hobby time to working as a reporter for the school's newspaper, The Stanford Daily as a nostalgic nod to a recently departed grandfather who was interested in journalism. Baker's first three stories for the paper on three different topics attracted exponentially larger responses from the campus community and the larger world. Baker just graduated from Stanford in the spring of 2026 and wrote How to Rule the World as an analysis of the culture at Stanford that has morphed far away from one centered on academic excellence and scientific integrity to one fixated on monetizing ideas into extreme wealth for not only students and faculty but for Stanford itself. The forces described in Baker's book meld seamlessly with other stories of corruption and bubbles, with the Artificial Intelligence bubble being the most obvious and ominous.

Before attempting to summarize the book and tie it to larger trends, it is worth simply stating that Baker's book is highly recommended and worth the time to read it. Universities were already showing signs of decay from "publish or perish" mantras that were already pervasive in the 1980s. Readers who attended such schools thirty plus years ago will have no difficulty recognizing those forces at work in Baker's book but will be astonished at how that pressure has grown exponentially more intense given the decision by many universities to promote a business mindset among their faculties and the millions of dollars available to faculty who are willing tear down the wall of separation between academia and crass commerce. Readers with children attending elite institutions like Stanford will gain a much clearer understanding of the pressures applied to students from literally their first day on campus and the level of farce associated with previously promoted ideals about the purpose of a college education.

Theo Baker doesn't necessarily attempt to cover this much ground in the book himself, but his experiences and analysis easily blend into larger themes being discussed across current media. To explain those ties, the book itself will be summarized then tied to other trends seen in academic fraud, venture / vulture capital and "innovation" and more general business fraud and regulatory failure.

A Review in Brief

In the larger scheme of things, Theo Baker is probably not a typical recent college graduate. He applied and was accepted at Stanford University. He was able to graduate from Stanford in 2026. He actually has two fairly famous parents, Susan Glasser who writes for The New Yorker and Peter Baker who has worked at The Washington Post, The New York Times and MS NOW. It seems fair to presume he thus came from some financial means. But in the context of Stanford, Theo Baker was likely a very middle of the road student on campus -- just an "average" student in a place where all of the children are far above average.

This "averageness" is important to emphasize for several reasons. First, all of the events described were initiated by a first-year student working for a student paper who came across the topics randomly. The topics didn't take any superhuman skills of discovery or analysis to describe. They simply required someone to see them, not look away and write plainly about them.

In a nutshell, Baker came to Stanford planning to major in computer science and pursue a technical career. He was NOT interested in a career related to his parents' journalism jobs though he respected them and those roles. However, within a few weeks of arriving for his first fall semester, he decided to join the school newspaper as a reporter. He was randomly assigned stories as events cropped up on campus and each of his first three stories attracted exponentially more attention -- both on-campus and off -- than the prior story.

Three Key Story Arcs

The first story, published October 24, 2022, was titled Inside "Stanford's War on Fun" and described how policies first imposed by the administration during COVID lockdowns had been expanded into a more general paranoia about ANY student gatherings due to alcohol abuse, date rape, etc. and had subsequently destroyed any semblance of typical college life on campus. Baker's piece, written after he attended a typical dud party and spoke with "safety officials" assigned to enforce the rules, punctured the absurdity of the extreme policies by pointing out how many students were simply doing their drinking AWAY from the campus adding to DWI dangers or drinking ALONE in their dorms without anyone to stop a spiral.

The second big Baker story seemed to be a fluke unrelated to overblown "War on Fun" concerns about "safety." The story, Stanford knew about the campus imposter for a year. He kept coming back. was published October 31, 2022 as a follow-up to a story on October 28 that involved an incident where a man was found to be masquerading as a student living in an empty dorm in the basement of one of the dorm buildings. In the October 28 story, campus officials described the incident as a one-off, nothing to see here. Baker did some follow-up that discovered the same man had engaged in the same poser scam MULTIPLE TIMES at Stanford. Those prior events had not been shared by officials nor mentioned when he was discovered again in 2022. This larger track record of an outside unknown party living inside dorm facilities punctured the university's prior attempts at preserving an image of an idyllic campus providing safety for all of its students, faculty and staff. Officials weren't just failing in their attempts, they were lying about prior incidents and memory-holing them.

Baker's third big story extended far beyond the Stanford campus in its reach. It started with an email from an alumni who had started following Baker's writing on the War on Fun. The alumni mentioned a comment in a blog post from a few years prior regarding concerns about several scientific papers involving drug research that were found to have duplicated photos included as multiple exhibits. Discovery of these types of frauds was becoming more common as more academic and scientific journals went online and made incidents of plagiarism and fraud easier to mechanically find. This tip was unique because one of the co-authors of the articles involved was Marc Tessier-Lavigne (MTL), the President of Stanford University.

The tip for the story arrived shortly after the imposter story but Baker did not immediately publish a story. He examined the original blog post with the allegation, reviewed the original articles at their original publication web site, compared the photos himself and sought technical counsel from established scientists. One called Baker back immediately, simply saying "Don't do this. Marc Tessier-Lavigne is unassailable, and you do not want to go after him." Over the next month, Baker contacted a microbiologist named Elisabeth Bik who had begun focusing on identifying scientific fraud specifically related to image fraud. A list of nine papers authored by MTL referenced in allegations was sent, she reviewed and found four had no real issues but the other five all had glaring signs of image manipulations, not to prettify the image for publication but to enhance positive observations or mask non-confirming observations in the article.

Based on that initial feedback, Baker wrote his piece, provided a copy to MTL for review prior to publication, was given nothing in response and published to the story on November 29, 2022, entitled Stanford president's research under investigation for scientific misconduct, University admits 'mistakes'. Between October 31, 2022 and December 31, 2023, Baker wrote a total of thirteen stories on the MTL saga that involved numerous concerns not only about MTL's integrity but that of the university administration as well.

• The fraud concerns over MTL's work pre-dated his hiring as president of Stanford and, per MTL, were communicated to the search committee by MTL yet the committee seemingly ignored them.
• In the rare cases where MTL responded to the paper's stories publicly, he did so using his official Stanford email address, presumably in an attempt to imply the university itself agreed with his claims and attempts to refute the allegations.
• The investigative team formed by the university administration refused to grant immunity to sources queried about the case.
• One participant in the investigation of MTL was forced to exit the review after it was found the participant held $18 million dollars in stock in a firm that MTL co-founded. The administration was too incompetent to identify such obvious conflicts of interest when forming the investigative panel in the first place.
• A 2009 paper co-authored by MTL during his employment at Genentech was cited in internal documents as foundational to Alzheimer products Genentech was developing which had triggered a pending buyout by Roche, a purchase that would net Genentech executives like MTL millions of dollars.

On July 19, 2023 MTL eventually resigned from his position as president effective August 31, 2023. Terms of his agreement required him to retract three papers and post clarifications for two more. By December 31 of 2023, MTL finally agreed along with his co-authors to retract the 2009 paper authored at Genentech, citing image anomalies and biostatistical errors while denying actual falsification of data.

Of course, it is crucial to note that MTL did not receive a SEVERANCE package from Stanford. He only relinquished his role as president. He is still employed as a faculty member and continuing his "work" in neuroscience. He even has his own sub-domain within stanford.edu for his own lab where he publishes his own responses like these

Addressing questions and mistaken claims about my research

The persistence of inaccurate claims: a brief commentary on recent reporting in The Free Press

to questions about the scandal as they continue to pop up. It isn't clear exactly how much work he is actually doing, especially since he formed a new startup company Xaira Therapeutics in April 2024 which claims to be developing AI technologies to more efficiently search for new drug therapies -- "making biology more computable" as the firm's own web site states.

Perversely, an article in the Stanford Review -- Why Stanford Hides Massive Executive Paychecks In Secret Contracts -- did provide some clarity on how much Stanford continues to pay MTL. In 2025, he was paid roughly $2 million dollars in salary, presumably as part of a deferred compensation plan offered to many Stanford leaders, essentially shifting earlier compensation from his term as president into subsequent years to smooth out tax burdens like any other corporate pay package.

The Stanford Inside Stanford

The MTL story arc probably occupies about sixty percent of the entire book. The balance of the book addresses a variety of patterns and behaviors that Baker terms "the Stanford inside Stanford." That term sardonically expresses the conclusion Baker reached within weeks of joining the Stanford community that the exclusivity of the elite school seen from the outside is a mere fraction of the exclusivity experienced WITHIN the university, because of its decades-long incestuous partnership with the Silicon Valley business environment and a conscious decision by university administrators to actively promote this mode of operation.

Baker doesn't quite set the stage for his observations this way but this is the essence of what takes place at Stanford (and presumably at least a few other elite universities).

Imagine you are a venture capitalist with $20 million dollars looking for the next "10-bagger". An investment that will return 10x your initial angel investment. In theory, "angel investors" have expertise in some line of business, follow developments in that line and attempt to find individuals or firms in those lines who are investigating promising areas with business promise. Angel investing has very low odds but can have very high (10x, right?) payoffs. An angel investor MIGHT choose to take their $20 million and find twenty different individuals or firms and place a $1 million bet on each. If 19 burn through their million dollars and fail, well, it's a tough business. If just ONE of the bets gets traction and at least advances to IPO, that angel might own a large enough stake to net a 1x or 5x payout. If the IPO goes REALLY well, the angel might get their ten-bagger. That still leaves them at a 50% loss over the entire $20 million. If one company turns into a hundred-bagger, now the angel has lost $19 million but made $100 million, netting $81 million.

Now instead imagine you are an angel investor who has not only been very fortunate but has gotten lazy or complacent. You could spend each year scouring the country looking for your twenty candidate investments for your next $20 million. That's a lot of work. But what if you are located on Sand Hill Road in Palo Alto, California and have coffee every morning on campus at Stanford which traditionally recruits students from the top 1% of the country and typically has dozens of students graduate, form new companies and make millions (at least for a while) on new biotechnology, new software product ideas or new hardware advances? Wouldn't it be a lot easier to just hang out at Stanford, suck up to the students and get them hitched to your wallet before any other investor has a chance to nab their ten-bagger? Come to think of it, a lot of these students don't seem to wait around to graduate before going into industry so you better start recruiting them early. How early? How about the first week of their freshman year?

THAT is the essence of the environment at Stanford described by Baker.

The book starts off with a glossary, providing tongue-in-cheek definitions of commonly heard terms on the Stanford campus. One term is wantapreneur, a term for a student who says they want to be a traditional entrepreneur who might take classes in accounting, financial management, marketing, etc. along with some core discipline to ensure they have a well rounded set of skills to operate a company from startup to established firm. At Stanford, wantapreneur is a term of derision. The top-tier students all profess to wanting to be builders because venture capitalists think they want geniuses who can spit out some quick idea that can be converted into a shell of a company to quickly shop it from angel phase to IPO to yield a quick cashout... For the angel investor... The angel investor cares nothing about the "genius" with the idea, they care nothing about the viability of the business past IPO or any of the investors in the IPO. They just want to turn a profit. (Theranos anyone?)

Another term Baker includes in his glossary is anti-signal, which describes a characteristic that, to the true insiders, means the opposite of what it means to outsiders. At Stanford, this moral-free philosophy and fixation on quick wealth is now common among students themselves and is invisibly re-enforced by a variety of secret organizations that quietly seek out those sharing the get-rich fixation and exclude students with a more grounded view of being an entrepreneur. The title of the book, How to Rule the World is actually drawn from the name of one such organization. Details are fuzzy but this organization was founded around 2018 by a junior at that time who might have originally been trying to offer an upper-classman's perspective to younger students on the realities of transitioning from student to entrepreneur and optimizing your wealth opportunities. However, that junior, identified as Justin Lewis-Weber, graduated in 2020 yet still operates this "club", requiring worthy new members to be chosen by prior members and screened by Lewis-Weber. Those selected not only take in the seminars but gain access to contacts for all prior selected members, offering a potentially lucrative set of connections.

Organizations like this may have started with more innocent aims and may have adopted such over-the-top names as a bit of college humor and sarcasm but the actual ideas now being pushed are notably more cynical and Machiavellian after multiple years. The leader of this particular "club" actually told Theo Baker "The only people who really understand the world are the literal children of billionaires." Well, so much for the rest of us.

The Larger Context

Baker's book focuses solely on his experiences as a student at Stanford and his reporting of issues involving Stanford but his observations reflect patterns of complacency and corruption that have spread throughout academia and business. These patterns have been the subject of multiple books over the past decade.

Self Preservation - The Only Mission That Counts

The book Bad City by reporter Paul Pringle summarized his multi-year effort on the staff of the Los Angeles Times to cover TWO stories involving the University of Southern California that eventually exposed corruption not only at the university but with the Pasadena Police Department and his own newspaper. The first story involved the dean of USC's medical school, Carmen Puliafitto, who was found to have paid to keep teen-aged girls in neighborhood motels, paid to keep them strung out on drugs and have used them for ongoing sex. His eventual outing started after one such girl overdosed in March of 2016 in a swanky hotel room rented in his name. He resigned three weeks later of his own accord but any public record of what had occured essentially vanished within the Pasadena PD the day of the incident.

The day after the death, Pringle received a tip from a worker at the hotel who described the scene, mentioned a man claiming to be a doctor at the scene, mentioned the man claiming that he was handling it and mentioned that Pasadena police ceded the scene to the "doctor" and didn't push public EMTs to command the scene. Pringle traced the room number and last name of Puliafitto, began investigating and spent over a year fighting with his own paper to publish an account of the people involved, what really happened and what didn't happen to save the victim. The Los Angeles Times management didn't want to piss off USC management, in part because USC is the single largest employer in the LA metro area and wields enormous economic clout in the region.

Pringle's second story involved a gynecologist George Tyndall who served on the staff of the USC campus medical facility for students. It was eventually discovered Tyndall was responsible for HUNDREDS of cases of rape and sexual abuse of patients and complaints of abuse, harassment and unprofessional conduct. Tyndall had been employed by USC since 1989 and complaints dated as far back as 1991 yet USC did nothing until another round of complaints from other staff in 2016 triggered internal investigations. USC eventually forced Tyndall to retire in 2017 and he was later arrested in 2018 by the LAPD. Again, Pringle's own paper was hesitant to publish the story because of USC's regional influence.

The unifying theme between the Pringle and Baker books is a pattern of large institutions becoming so inwardly focused to protect their own reputation and perceived interests that they become toxic to their surrounding communities. Beyond a certain size, nearly every institution - academic, corporate, charitable, or social -- becomes administratively warped to the point where self-preservation and perpetuation becomes the most important goal.

Cornering a Monopoly on Innovation

Prior commentaries on this blog regarding a Grand Unifying Theory on Creativity, Productivity and Specialization addressed trends over the past few decades with "innovation" in corporate settings. CEOs like to think as "innovation" as just another knob on their dashboard that can be cranked up or down, instantaneously, at will to solve a competitive problem or cut costs to juice quarterly profits to meet a bonus goal. At the core of that mindset is a fallacy that valuable ideas are just another inventory item waiting to be purchased on a just-in-time basis. They are most assuredly NOT.

Baker's anecdotes of student life at Stanford go beyond the absurdity of supposedly knowledgeable investors showering a freshman or sophomore student operating a real business out of their dorm (think Michael Dell physically assembling and shipping PCs from his UT dorm...) with millions of dollars. At Stanford, outside investors are so CONFIDENT one of these nerds is going to found the next unicorn firm to IPO at one billion dollars, they now offer "pre-idea" contracts to freshman students within WEEKS of starting school. Investors are attempting to lock up "talent" before that "talent" has even completed a single class project or taken a mid-term exam.

What Is the Actual College Experience in 2026?

For some readers of Baker's book, the biggest shock won't be the scientific fraud on the part of a man who served as university president. It won't be the insular, discriminatory secret club culture that creates a divided class polarized between people who are already rich and plan to become far richer by any means available versus those with a more humanitarian approach to their career and academic training. No, the biggest shock will be the contrast between the stereotype of a typical student's daily workload and attitudes towards that work at an elite college and the reality.

In the portion of narrative not explicitly tied to the MTL case and the macro observations of investor meddling in student life, Baker describes blowing off multiple large homework assignments in a core class for his would-be major and waiting until one or two days before due date to begin work. He describes deferring school work out of an entire WEEK to participate in an extracurricular "club" conducting a hackathon for high school students. Reading about his involvement in TreeHacks just seemed odd. What would a typical first year CS student have learned approximately seventy percent through their first year of CS coursework that would be worthy as a foundation to teach high school students?

Forty years ago, yer humble obedient scrivner WTH attended engineering school at what was thought to be at the time a "top twenty five" school (per US News & World Report). Admissions were assumed at the time to come from the top 10% of students nationwide. After the first year of mind-numbing core classes in calculus, differential equations and physics, I found myself falling from the top 1% of a high school class of about 400 to the fiftieth percentile of a class of about 250 EE, CS, MS, SSM and CE engineers. One of my closest undergrad classmates was arguably the smartest kid in the entire class. While double majoring in electrical engineering and biology to prep for medical school, he didn't just get "As", he didn't earn a single A minus in four years. He was simply brilliant at all of the advanced mathematics that really constitutes the core curriculum for most engineering majors. But being preternaturally comfortable with all of those mathematical manipulations and transformations was not the same as being able to complete the work in twenty or thirty minutes. A single problem on a homework assignment might still require four or five PAGES of hand-written integrations, etc. And my 4.0 friend was right there with the rest of us less gifted mathematicians working until 1am on due date to complete assignments.

Maybe it's possible that (relative) rubes like me who DIDN'T get into Stanford cannot comprehend the mind-numbing brilliance of the students who ARE admitted to Stanford. Maybe they ARE so gifted that they can comfortably blow off problem sets until 11pm the night before due date and knock them out in fifty minutes without breaking a sweat. Maybe they've been writing native C-language software drivers for an ASIC based control circuit they designed when they were fourteen years old in junior high and aren't challenged by a homework project in an intro computer science course illustrating the fundamental concepts of a finite state machine...

...or...

...Maybe it is possible that the cultivated sense of entitlement is so strong at Stanford and other "elite" colleges that expectations of students to finish coursework have slumped. Maybe grading curves have softened realities so much that a student can skip a couple of homework projects and still pull out a B minus grade and still look like they earned a B minus in an incredibly difficult curriculum and still merit respect and consideration by future employers...

...Or maybe it is possible that professors have grown so complacent with teaching and alternative web content covering the same material is so much better that students DON'T really have to attend classes, can learn on their own from the web and still complete the coursework and pass the tests.

Regardless of which of these scenarios might reflect some degree of truth, none of the them reflect well on the true value being provided by such institutions. EXCEPT for the "connections" made possible by simply being an alumni.

Speculation + Complacency = Catastrophe

Baker's observations may have been shocking to him as a new freshman at Stanford in their kind, but for anyone else watching Corporate America and elite universities evolve over the last twenty years or so, Baker's observations are only shocking in degree. No one should be surprised that a "scientist" and "professor" who routinely alternates between academic roles and business ventures would struggle with a decision to retract his own scientific publication if it meant tanking his own stock holdings by fifty percent (even if still leaving him with tens of millions of dollars). No one should be surprised that venture capitalists have blown off doing their own homework when picking new business ideas and instead are simply trying to corner the market by locking up talent before their college email address is even set up. And no one should be surprised that students who were raised amid a get-rich-quick culture and pressured from the crib to get into the best prep schools to grease the skids for getting into an elite school would arrive at that school fixated primarily on converting that opportunity into money.

The dangers posed by the present bubble in Artificial Intelligence aren't even an open secret. Thousands of reports and opinion pieces have been written describing the circular revenue being touted by participants. Stories are already being filed about nearly $500 million construction bills dating back to 2024 going unpaid by SpaceXAI that reflect cash crunches are already taking root. As of August 2026, a top trending story involves a 24 year-old named Leopold Aschenbrenner who built a position worth $48 billion dollars which collapsed to ZERO in THREE DAYS.

Aschenbrenner worked briefly at FTX after graduating from Columbia University in 2021, resigned the day FTX collapsed in November of 2022, took up a job at OpenAI in 2023 and was eventually fired in 2024 after communicating directly to its board about security concerns regarding the firm's intellectual property. He then wrote a blog post identifying investments in AI infrastructure firms as the highest return, lowest risk approach for profiting from the bubble, regardless of whether AI itself eventually paid off as a technology. This was actually a very SOUND investment strategy but the 24-year old, with ZERO prior experience in Finance and ZERO experience operating a hedge fund, actually created a hedge fund around this premise. His blog post was shared millions of times and attracted attention from actual hedge fund investors who saw him as a typical serial genius (but go back and look at that employment record...) and showered him with tens of millions of dollars which he promptly levered into vastly larger and more complicated positions he didn't understand. The dollar size of his bets failed to account for the lack of liquidity involved with such high leverage. It's one thing to devise a hedge strategy with only one or two million dollars in a position. It's another thing entirely to have a few BILLION tied up in something that might trigger a margin call requiring BILLIONS to be sold when the market isn't interested. It only took a couple of days of alignment to trigger margin calls big enough to evaporate a $48 billion dollar portfolio. Ironically, the name of his hedge fund? Situation Awareness.

Situational awareness, indeed. It seems like a long-forgotten concept. It certainly seems forgotten at many elite universities who are not only failing to teach skills required to recognize these ethical and economic dangers but are indoctrinating students into the corrupt practices and attitudes producing them.

WTH

The CBS Evening News with Byron Allen

Media in 2026 has been replete with stories of hiring decisions, firing decisions, content battles, cancellations, plummeting ratings and plain amateurish / incompetent execution of basic news groundwork. Reactions IN the media about these media stories actually pose their own concerns because those reactions frame the play-by-play in the context of assumptions about how media (news media in particular) SHOULD operate within a corporate / capitalistic framework and the goals of the owners of the corporations controlling them. The concern stems from the fact that few explain the assumptions being made and the arguable fact that none of those assumptions are remotely true in the current environment.

Golden Era Assumptions

News of conflict and failure at CBS seems to be arriving nearly continuously. Paramount offering a $16 million dollar settlement in a frivolous lawsuit filed by Trump against 60 Minutes editing of an interview as a carrot to gain approval to be bought by David Ellison and Skydance. CBS canceling its top-rated late night program claiming it was losing $40 million per year. CBS placing "independent media" op ed writer Bari Weiss in charge of the entire CBS News division. 60 Minutes delaying a story on illegal immigrant detention centers. Weiss pressuring 60 Minutes producers to adopt rules allowing story subjects to pick their preferred 60 Minutes anchor. CBS firing Sharon Alfonse. CBS firing Scott Pelley. CBS Evening News ratings tanking. Evening News anchor Tony Dekoupil having to report on Trump's May 2026 China trip from Taiwan because no one left on the Evening News staff knows how to arrange logistics for an overseas news trip. CBS shuttering its CBS Radio operation in place since 1927. Remaining 60 Minutes anchors huddling privately to discuss whether there's anything to return to in its next season.

As outsiders looking in on this chaos add play-by-play analysis, that analysis tends to fall into certain ruts based on a consistent set of assumptions about the actual goals of those in charge of CBS and CBS News in particular. Most commentary seems to focus on scoring the choice of tactic aimed at achieving goal X or the quality of the execution of that tactic towards goal X. Virtually no commentary is addressing the elephant in the room... Is X even still a goal for this company?

The assumptions driving how participants and critics frame these debates date from older nostalgic understandings of the balance of power between a "news" organization and any larger business parent happening to own that news organization. In hindsight, this nostalgic understanding of the way things use to operate was never 100% true even in the good old days. Most definitely, those assumptions are demonstrably false in the current environment.

What are the assumptions being made?

• News organizations inside corporate entities still enjoy a magical protective bubble stemming from a quaint sense of noblesse oblige on the part of executives. This bubble somehow ensures story selection and editing will never be tainted by crass concerns about profits or fears of offending powerful business, political or social figures. Those running the news would always know what the right thing to do was and would always have the freedom to do it.
• Corporate owners of media outlets are noble enough to view the cost of news operations as a "loss leader" or a means of burnishing a larger corporate image that provides value beyond the bottom line of the news organization on its own profit and loss statement.
• News organizations should strive to be defensibly non-partisan, providing timely information on events of equal importance on any side of a contested topic.
• Even if topic selection and content editing won't be purely unbiased in any particular direction, it will still be predominately fact-based.
• When conflicts arise between mere business interests and news interests, news interests should take priority.
• Business decisions about entertainment content do not have to be proactively "balanced" according to some perceived scale of bias. If a show captures viewers that seems to be positioned at point 0.25 on the 0 to 1 scale, the media owner isn't required to air a program with content positioned at 0.75 to "balance" out the first. If they can find such a program and viewers tune it in, they can certainly air the program but they're not REQUIRED to air it.

The reality is that these assumptions are not only demonstrably false in the context of CBS and its new parent conglomerate, these assumptions are no longer true for any large media outlet. It is ipso facto the case that any large media conglomerate that includes "news" entities within it is already tainted by the forces applied by boards and shareholders on any sufficiently large corporation.

In December of 2024, ABC settled a "defamation" lawsuit filed by Donald Trump the citizen in March 2024 that no legal expert in the country thought required settling prior to trial. Why? Because Trump won the 2024 election and Disney -- ABC's owner -- didn't want to start off the next four years on the new President's shit list.

Comcast sold off a variety of cable channels including MSNBC and CNBC, in part because viewership has shrunk in lockstep with cable-cutting of all video subscribers in cable / satellite TV. However, MSNBC and CNBC had relatively good viewership in sought-after demographics but it seems apparent Comcast felt those economics weren't worth the cost of content on MSNBC generating daily threats of retribution against Comcast's larger interests from a thin-skinned President.

Who's Paying Whom?

The programming strategy behind the elimination of The Late Show With Steven Colbert and replacing it with the Byron Allen show Comics Unleashed demonstrates many of the dynamics at work with "media" in general and "linear television" in particular at this point in time. (Linear television refers to programming delivered to viewers at a fixed schedule rather than "on demand" as saved content that can be played / paused / rewound / fast-forwarded / skipped as part of its core delivery experience).

Not all programming appearing on linear television channels is produced and distributed with the same financial goals. Think back to NBC content in the 1980s and 1990s. For nearly twenty years, NBC succeeded at identifying and contracting with a series of creators who delivered sit-com and drama content such as Cheers, Seinfeld, ER, Friends, Frasier, Mad About You, The West Wing, etc. that triggered a virtuous circle of wealth for all involved:

• NBC paid good money to a writer / producer for a concept and show scripts and the production of the show
• the content won an audience and advertisers clamored to reach that audience by paying NBC more money for ad slots
• NBC made more money allowing more speculation on more writers / producers to find the next hit show
• NBC could schedule new shows after existing hits to accelerate the adoption cycle for new shows, making them the next big hit
• many viewers got to the point where the consistency in programming became its own brand ("Must See TV" on Thursdays), further helping viewership and lead-in ad revenue at local stations before and after prime-time blocks

Did NBC itself create these shows or own them? No. Paramount produced Cheers and Frasier. Sony produced Seinfeld and Mad About You. Warner Brothers produced ER, Friends and The West Wing. NBC owned time slots during these shows and made its money by paying the creator $X million for the right to air the show on its first runs for Y years prior to syndication while collecting substantially more than $X million in ad revenue, turning a profit.

When most people think of how "television" works as a business, that's the model they imagine at work.

But that's not the only way content makes onto a television channel (either broadcast or cable). The opposite extreme is easiest to explain by thinking of your local television station and your local creepy Christian mega-church pastor. For some communities too small to sustain the appetite of a local holy roller, think of some of the national charlatans like Jim Bakker, Robert Schuller or Joel Osteen, who did / do the same thing across multiple markets. What do they do?

They buy ALL of the ad slots within a given time slot from a local television station (typically outside of prime time hours - often early Sunday mornings). Rather than the station having to find content and pay for it for that time slot, the "church" provides the content. The "church" pays for all of the production costs. The local station just connects VIDEO IN from the megachurch to SIGNAL OUT and collects the money. The station really doesn't CARE if anyone watches. At most, the only thing the local station cares about is that the content delivered by the "church" isn't SO blatantly offensive to local mores that the content triggers viewers to avoid OTHER shows the station airs that WOULD reduce the ad revenue collected from other local businesses selling Chevrolets and appliances. (This model is also popular with sub-prime used car dealers.)

Until the last ten or twenty years of media consolidation, it would have been safe to say that no local station SOUGHT OUT a local mega-church or even a more traditional church and actively ASKED to place video crews to record services and broadcast them on local TV, either as a money-making ploy or as "public interest" programming for the local community. It would have been a safe bet to assume every one of these arrangements involved the church buying the time slot entirely and incurring all of the live production costs as means of getting its message out.

In the last ten or twenty years of media consolidation, it is possible that some of these conglomerates such as NextStar (owner of more than 200 stations), Gray Media (owner of 113 stations) and Sinclair Broadcaster (owner of 193 stations) might find philosophical synergy with mega-church content and might apply some pressure to local properties to cut deals to air such shows, altering the financial balance somewhat. It's definitely already the case that some of these conglomerates (Sinclair specifically) supply pre-recorded "must-run" content to local properties who air them during local newscasts. These segments are structured and produced to meld with regular reports but present grossly distorted explanations of basic political and constitutional principles.

What's new in the last year is that this "mega-church" production model to provide content to fill a time slot and increase profits for stations is now being adopted by the networks directly. The change at CBS to dump Steven Colbert for Comics Unleashed is the first notable example. The show Comics Unleashed is produced by Allen Media Group, a parent company owned by Byron Allen, who many might remember as one of a collection of hosts on NBC's Real People show of the late 1970s. The show WAS popular... Initially... The show lasted five years then tanked, leaving only a memory of what most people today view as quintessential bad 1970s television. Right up there with the infamous "Roller Disco" episode of CHiPs.

Byron Allen moved onto other ventures, starting with a concept of reselling celebrity interviews collected during press junkets for new movies, a concept which had only been adopted by a few hundred DJs at radio stations across the country who attended the same junkets to create "content" to fill morning drive-time on the radio. The content was completely generic and low quality but cheap to produce.

Most radio stations and TV stations abandoned the "celebrity gravy train" model for "content" by the early 2000s but Allen created the Comics Unleashed program in 2006 using the same formula:

• The "talent" that appears is typically only paid union scale wages -- about $1000 currently
• Talent that appears is NOT paid any residual or royalty for subsequent airings of their appearance, which are frequent and may continue appearing for years, potentially dulling the comedian's reputation with new fans
• The comics bring their own material so there's little if any fixed expense for staff writers
• Single-set physical logistics with common camera views -- little reliance on camera operators and editors for "live" production, everything is edited and assembled into a final form after the fact, days ahead of airing, further lowering production costs

The Comics Unleashed show is a perfect example of this low-cost, assembly-line "content" model. Most current CBS viewers might have assumed that Byron Allen would be delivering NEW episodes of Comics Unleashed after taking over the Late Show slot. Not exactly. Allen is expected to product 132 new half-hour episodes the first year (enough for 50% of the weekdays per year) and the second half-hour will re-run old shows recorded between 2006 and 2016. It's not clear if production will ramp up to provide more new content in coming periods.

That's essentially saying only twenty five percent of the content airing in that weekday hour-long time slot will be "new." But few willing to watch will likely notice any difference between the "old" and "new" because a key tenet of the content generation model for Comics Unleashed is to avoid ANYTHING remotely current in the material. This approach has the obvious benefit of increasing the shelf life of the content produced but in the current environment, that has the additional benefit of assuring any content won't touch on anything controversial that might offend the corporation owning the network or those it is trying to suck up to.

Byron Allen's larger conglomerate Allen Media Group owns a collection of cable / satellite TV channels all following a similar model: The Weather Channel, Comedy TV, Cars TV, Pets TV, Justice Central, etc. Either limited production costs with a very unchanging format (weather) or recycled content from other sources "curated" into "fresh content" merely by being lumped together with an airtime schedule. And this type of content is seen by viewers for exactly what it is -- completely bland, forgettable content that kills brain cells through mere contact.

The fact that this content provides little draw to customers to continue cable or satellite subscriptions is very evident to those providers which is why they have been unwilling to pay large premiums to carry the channels. Allen actually leveraged that against Comcast and Charter by suing them for racial discrimination and violation of the Civil Rights Law of 1866 by "refusing to make contracts" with his firm because Allen is African American. (Nooooo... we're unwilling to pay $X per subscriber to carry your channels when we have exact viewership data showing a tenth of a percentage point of our subscribers WATCH these channels when we carry them...) The suits were filed in 2015 and went all the way to the US Supreme Court which tossed out his case in 2020 with a rare, unanimous 9-0 decision. Allen later settled the suits privately with each provider agreeing to continue carrying some mix of his channels for undisclosed amounts.

Late Show Production Economics

If you believe comments from CBS used to justify their cancellation of Steven Colbert's show and the selection of Comics Unleashed to replace it, the economics of Allen's business model seem to make a decision to ditch Colbert obvious. CBS claims yearly production costs for Colbert resulted in a net loss to CBS of roughly $40 million per year. In contrast, because Allen is buying up the air time and producing the show on his own dime, CBS zeros out all production costs and collects about $15 million from Allen, producing a "swing" of $55 million from a $40 million loss to a $15 million profit.

These numbers tossed out by CBS seem, to say the least, quite suspect. Colbert's most recent contract paid him $15 million per year. CBS claimed yearly ad revenue for the show was around $60 to $70 million (down from $120 million earlier in the run). To lose $40 million per year, that would require production costs of nearly $100 million, which, after Colbert's $15 million salary, leaves $85 million for the 200 staffers collectively -- an average salary of $425,000. Clearly, camera operators, gaffers, ushers, teamsters working the stage, etc. were not making $425,000 yearly so this is not an accurate means of reverse engineering the real labor costs. Even the writers likely capped out around $200,000 plus additional pay for on-air skits, etc. Guests are likely comped with luxury hotels and transportation so with 2 "couch guests" and a musical guest with four band mebers per night for 162 shows per year, the lodging alone comes to about $583,000 -- a pretty inconsequential cost in the big picture.

CBS purchased the Ed Sullivan Theater in 1993 for about $4 million and spent about $4.5 million renovating it for the Letterman era. One can image renovation costs in 2015 were three times that or roughly $13 million. CBS also captured a tax abatement of $5 million from New York City to keep the show in the city. One would assume CBS leased the space to Colbert's production company so real estate costs are already factored into this $100 million paid to his production company.

Of course, missing in discussions of the profitability of Colbert's Late Show or its equivalents is any mention of the obvious purpose of these shows to begin with. They are NOT intended as a means of paying a handsome raconteur to entertain the masses. If that happens, that's okay but that is NOT the goal. They are not necessarily required to turn a profit on their own, though if that happens, that's a plus. These shows have existed from their inception in the 1960s as promotional vehicles to use in flogging the latest offerings from TV networks, movie studios and publishing houses. As long as these shows are pulling in three million viewers per show, that's three million consumers seeing promos for upcoming movies, albums and books with ownership stakes benefiting...? These same corporate owners. None of those intra-corporate revenue synergies are being reflected in the suspect accounting of the "profits" from these shows.

So What's Missing in the Analysis?

So if all of the old assumptions about how news operations should operate are false, how does it affect coverage of the latest strife? As an example, one theme in stories about CBS News is that Bari Weiss is the WORST person who could have been selected to run the organization, even if one is willing to concede CBS News had issues and needed to change. This line of thought identifies these problems:

• Weiss' only experience is as an individual reporter and op-ed writer, not an editor or TV production executive or "line of business" executive.
• Weiss' choice of "stars" may actually drive current viewers away, hastening the collapse in ratings.
• Weiss' pursuit of more conservative content to appear on CBS outlets won't attract loyal conservative viewers of more right-wing channels.

All of these stated concerns about Weiss' tenure to date all assume the goals of operating a news organization and consistently airing "fair" content still remain. None of these assumptions can be proven with certainty at CBS. It's not clear they can be proven with certainty at Comcast for NBC or Disney for ABC. These corporate owners may not feel ANY obligation to keep a news organization running. Unlike local licensing rules for broadcasters, there are no FCC mandates applied to national network operators to provide recurring news shows.

The takeaway is that the current corporate owners of top "news" organizations in the United States respect no boundaries between editorial decisions within news teams and corporate financial goals. If executives conclude it will improve profits over the next three years to replace an independent news organization with pre-fab content assembled by 20-somethings who learned how to use DaVinci Resolve while running a YouTube channel but know nothing about history, economics, science or civics, they will do it in a heartbeat, even if the country loses all visibility into what the government and courts are conspiring to do to surrender control of society to our oligarchs. There is no assurance that The CBS Evening News with Byron Allen or something very similar to that model isn't already being pitched to executives at Paramount (or Comcast or Disney).

If there's no FCC mandate at the network level to produce and air recurring news shows on a daily basis, what's to stop existing national networks from abandoning such efforts? Absolutely nothing other than habit and unverified, unspoken assumptions that such an alternate universe somehow cannot exist. Such a universe absolutely CAN exist. The difference between a world with a thirty minute The CBS Evening News with Walter Cronkite show existing and a world with The CBS Evening News with Byron Allen or a world with no CBS news at all is the difference between having a William Paley at the helm versus David Ellison. Paley did not enjoy a perfect record on his journalistic independence scorecard but his overall management arc yielded "the Tiffany Network." Ellison in contrast has arguably trashed any semblance of that network still standing in a matter of months, and not by accident.

WTH