Socratic Seminars: GREED NATION & POISONED FOR PROFIT Reading Passages

 About This Socratic Seminar Series of Reading Passages

Are We Teaching Kids to Think Critically — Are We Giving Them Anything Real to Think About?

Here's a question worth sitting with:

A medication costs $4.73 to manufacture. The company charges $935.77 for it.

Not $50. Not $200. Nine hundred and thirty-five dollars.

That's not speculation. That's from a peer-reviewed study published in JAMA Network Open by researchers at Yale University and Doctors Without Borders. The drug is Ozempic. The company is Novo Nordisk. The profit margin is real, documented, and — here's the part that should bother all of us — completely legal.

This is the kind of fact that doesn't fit neatly into a textbook. It doesn't have a clean takeaway. It demands that students — and adults — sit with a genuinely uncomfortable question: At what point does profit become extraction? And who decides?

That's the question at the heart of two free reading series I've developed for high school Socratic seminars.

---

What These Reading Series Are

Greed Nation and Poisoned for Profit are collections of close-reading passages built entirely on verified, primary-source facts — peer-reviewed research, Senate investigation reports, internal corporate documents released through litigation, and investigative journalism from major news organizations.

All facts in this series are drawn from peer-reviewed research, official government reports, and reporting from major news organizations. Each passage includes primary sources students can verify independently. Students and teachers are encouraged to check every claim, seek out counterarguments, and form their own conclusions. Critical thinking — not agreement with any particular viewpoint — is the goal.

SERIES 1 — GREED NATION: When Profit Becomes More Important Than People

Four reading passages for high school Socratic seminar. Subjects: corporate greed, pharmaceutical pricing, housing exploitation, economic bubbles, and regulatory capture. Each passage includes 6 Socratic questions, 4 Food for Thought prompts, and verified primary sources.

1	"The $936 Pill" — When Medicine Becomes a Luxury A detailed examination of Ozempic's production cost ($0.89–$4.73/month) vs. its U.S. retail price ($935.77/month), using the 2024 Yale/JAMA Network Open study. Covers pharmaceutical patents as monopolies, the U.S.–Europe price gap (13x), Novo Nordisk's Senate testimony, and who goes without. Ideal for economics, health science, civics, and ethics units.
2	"Rusted Wheels for $2,500 a Month" — Vanlords and the Housing Crisis Documents the California vanlord phenomenon: property owners renting inoperable, unregistered RVs to unhoused residents for hundreds to thousands of dollars per month with zero tenant protections. Covers Bay Area housing economics, the regulatory gap that makes vanlording legal, policy capture in housing regulation, and the limits of market solutions to human need. Strong for civics, economics, and social justice.
3	"Boom, Bust, Bailout, Repeat" — Economic Bubbles and Who Pays Traces the shared anatomy of three U.S. economic collapses: the dot-com crash (2000, ~$6.2T in household wealth destroyed), the 2008 housing bubble ($700B bailout, 9 million jobs lost), and the early indicators of the AI investment surge. Explains mortgage-backed securities, regulatory capture by the financial industry, and the asymmetric distribution of profits and losses. Strong for economics, history, and financial literacy.
4	"Greed Nation: Connecting the Dots" — Synthesis A synthesis passage identifying the five-step pattern of corporate extraction visible across all three preceding passages: monopoly position → maximum extraction → policy capture → crisis → public pays. Addresses complicity, systemic thinking, and what changes would be required to alter the pattern. Designed as a culminating discussion for the full series, or as a standalone critical thinking exercise.

 

SERIES 2 — POISONED FOR PROFIT: How Greed Is Destroying Our Environment and Our Health

Five reading passages for high school Socratic seminar. Subjects: EPA dismantling, PFAS contamination, pesticide regulatory capture, fossil fuel climate denial, and environmental justice. Each passage includes 6 Socratic questions, 4 Food for Thought prompts, and verified primary sources.

1	"Dismantling the EPA" — When the Watchdog Is Defunded Documents the 2025 EPA budget and staffing cuts (proposed 55% budget reduction; 1,274 positions eliminated; all Environmental Justice programs ended), the reorganization of the Office of Research and Development, and the effort to reverse the 2009 Endangerment Finding that provides the legal basis for all U.S. climate regulation. Connects each cut to the industries that lobbied for it. Strong for civics, environmental science, and political science.
2	"The Chemicals That Never Leave" — PFAS and Corporate Cover-Up Examines the PFAS 'forever chemical' crisis: 12,000+ synthetic compounds, detectable in ~45% of U.S. tap water and in virtually every American's blood. Covers the 2025 NIH-funded USC study estimating 6,864 annual cancer cases from PFAS in drinking water; the 3M and DuPont internal document evidence of decades-long health risk concealment; and the 2025 partial rollback of the EPA's 2024 drinking water standard. Strong for environmental science, health science, and ethics.
3	"Banned Everywhere Else" — Pesticide Industry and Regulatory Capture Examines atrazine (banned in EU, still used on 60M+ U.S. acres; classified probable carcinogen by WHO in November 2025) and glyphosate/Roundup (Bayer facing 170,000+ lawsuits; $10B+ in settlements). Documents Monsanto's ghostwriting of academic papers, the 50+ private EPA-Syngenta meetings during atrazine review, the revolving door between the American Chemistry Council and EPA leadership, and the 322M lb annual U.S. use of EU-banned pesticides. Strong for science, ethics, and media literacy.
4	"The Longest Lie" — Fossil Fuels and 50 Years of Climate Denial Traces Exxon's 1977 internal climate research (which accurately predicted observed warming), the subsequent decades-long industry disinformation campaign, the IMF's estimate of $757B in annual U.S. fossil fuel subsidies, and the 2025 EPA rollbacks of vehicle and power plant emissions standards. Addresses environmental justice: air pollution's disproportionate impact on low-income communities and communities of color. Strong for environmental science, history, political science, and ethics.
5	"Connecting the Poison" — Synthesis and the System Identifies the six-step corporate deception playbook visible across all four preceding passages; examines how the costs of environmental harm are consistently shifted to communities with the least political power; challenges students to evaluate whether democracy can reliably produce protective environmental policy when the regulated industries have disproportionate access to regulators, policymakers, and media. Culminating discussion passage for the full series.

GREED NATION

A Socratic Seminar Reading Series for High School Students

When Profit Becomes More Important Than People 

This reading series examines one of the most persistent and powerful forces shaping American life: the drive for profit at the expense of human need. From the price of a life-saving medication to the collapse of the global economy, the readings that follow trace how unchecked greed has shaped—and repeatedly broken—the world we live in.

These are not opinions. They are documented facts, supported by studies published in major scientific journals, investigations by the United States Senate, and reporting by some of the most respected news organizations in the world. Read carefully. Think critically. Be willing to be uncomfortable. The questions raised here do not have easy answers—but they are questions every generation must face.

How to Use This Series

Each passage includes: (1) a reading built on verified facts, (2) Socratic Seminar discussion questions designed to push your thinking, (3) Food for Thought — deeper provocations to take the conversation further, and (4) Sources to Verify — so you can check every claim yourself.

 

PASSAGE 1  |  The $936 Pill

When Medicine Becomes a Luxury: The Cost of Staying Alive in America

The Numbers Don't Lie

In March 2024, researchers from Yale University, King's College Hospital in London, and the nonprofit Doctors Without Borders published a study in the prestigious medical journal JAMA Network Open. Their finding was stunning, and it was not disputed: the diabetes and weight-loss medication Ozempic — made by the Danish pharmaceutical company Novo Nordisk — can be manufactured for as little as 89 cents per month. Even including a profit margin and all production costs — the chemical ingredient semaglutide, the disposable pen device, the filling and packaging — the most the drug should cost to make is $4.73 for a one-month supply.

Novo Nordisk charges Americans $935.77 per month for the same drug, before insurance and rebates.

To put that in plain math: for every dollar it costs to make Ozempic, Novo Nordisk charges approximately $200 to $1,000. That is not a business model. That is a ransom.

The active ingredient — semaglutide — costs about 29 cents per monthly dose. The injection pen costs $2.83. Filling and packaging costs 20 cents. Other chemicals cost 15 cents. Add it all up: you get under $5. The company charges nearly $1,000. The difference — nearly $930 per patient per month — is pure profit, extracted from people who need this medication to manage life-threatening diseases including Type 2 diabetes and obesity-related heart disease.

A Drug the World Needs, at a Price It Cannot Afford

Ozempic is not an obscure medication. As of 2023, an estimated 2% of all American adults who visited a doctor had been prescribed Ozempic or its sister drug Wegovy. More than 10 million Americans are currently taking GLP-1 medications. For many people with Type 2 diabetes or severe obesity, these drugs are not optional lifestyle choices — they are medical necessities. People who cannot afford them face amputations, blindness, kidney failure, and death.

The company's own CEO, Lars Fruergaard Jørgensen, was called before the U.S. Senate Health Committee, chaired by Senator Bernie Sanders, to explain why Americans pay so much. The answer became clear when committee investigators compared prices around the world: Americans pay $935 per month for Ozempic while patients in Germany pay the equivalent of about $140, patients in Denmark (Novo Nordisk's home country) pay $186, and patients in the United Kingdom pay $92. The same drug. The same company. A price difference of more than 1,300 percent — in favor of every other wealthy nation except the United States.

Novo Nordisk earned more than $10 billion in just the third quarter of 2023 — more than half of its revenue from these two drugs alone. In 2023, it became the most valuable publicly traded company in Europe, surpassing the luxury goods giant LVMH.

The Patent as a Weapon

Novo Nordisk holds a U.S. patent on semaglutide that does not expire until 2032. This patent — granted by the U.S. government — gives the company the legal right to be the only seller of Ozempic in America for the next seven or more years. No generic manufacturer can produce a cheaper version. No competition is permitted. Meanwhile, Senator Sanders revealed that executives from generic drug companies told him they could produce and sell an identical drug for less than $100 per month — less than 10 percent of what Novo Nordisk currently charges — and still make a profit.

This is not a failure of the market. It is a designed absence of one. The company lobbies for patent protections. The government grants them. The patient — often a lower-income diabetic who cannot afford $12,000 per year — is the one left without options.

Novo Nordisk's Defense

Novo Nordisk argues that its high U.S. prices subsidize the billions it spends on research and development — nearly $5 billion in 2023 alone. The company also notes that 75% of its gross revenue goes to rebates and discounts to insurance companies and pharmacy benefit managers, and that patients with private insurance can access the drug for as little as $25 per month through the company's savings card.

Critics, including the Yale researchers, push back: the savings card does not help uninsured patients, Medicare recipients, or Medicaid patients. The billions in R&D spending — while real — cannot fully explain why the same drug costs thirteen times more in America than in other rich countries. And the profits accumulated by Novo Nordisk during this period reached levels that even the company's own CEO admitted were unexpected.

 

🔍 Socratic Seminar Questions

1. Is a company morally obligated to make a life-saving drug affordable, or does it have the right to charge whatever the market allows? Who decides?

2. Novo Nordisk argues it needs high prices to fund research for future drugs. Is this a valid argument? How would you verify it?

3. The U.S. government grants pharmaceutical patents that block competition. Should the government also have the power to cap prices? Why or why not?

4. Americans pay 13 times more than British patients for the same drug. What does this tell us about how the American healthcare system works — and for whom?

5. If you were a Type 2 diabetic without insurance, what options would you have? What does your answer tell you about the system?

6. Can a business practice be both legal and wrong? Give an example from this passage or from your own experience.

💭 Food for Thought

• The researchers who published this study said their goal was to 'have receipts' — to make the math transparent so no one could deny it. Why might powerful corporations prefer that this kind of math stay hidden?

• The drug industry claims it needs patents and high prices to fund innovation. But many of the original scientific breakthroughs behind GLP-1 drugs were funded by public universities and government research grants — meaning taxpayers already paid once. Should they have to pay again?

• Novo Nordisk is a Danish company. Denmark has universal healthcare and negotiates drug prices centrally. Its own citizens pay a fraction of what Americans pay for its drug. Is this ironic? Is it wrong?

• Martin Shkreli became infamous for raising the price of a lifesaving drug 5,000% overnight. Public fury was enormous. But Ozempic's markup over production cost is comparable. Why do you think the public reaction has been different?

📚 Verify the Facts — Sources to Check

Barber, M. et al. (2024). "Estimated Sustainable Cost-Based Prices for Diabetes Medicines." JAMA Network Open. — The Yale/Doctors Without Borders study on Ozempic production cost.

CNBC (March 27, 2024): "Novo Nordisk's $1,000 diabetes drug Ozempic can be made for less than $5 a month, study suggests." cnbc.com

NBC News (Sept. 18, 2024): "Bernie Sanders says Ozempic can be produced for less than $100 a month." nbcnews.com

Fortune (March 28, 2024): "Ozempic maker Novo Nordisk facing pressure as study finds $1,000 appetite suppressant can be made for just $5." fortune.com

U.S. Senate HELP Committee Report (2024) — Investigation into Novo Nordisk pricing. Available via senate.gov

 

PASSAGE 2  |  Vanlords

Rusted Wheels for $2,500 a Month: The Profiteering of Desperation

When Your Only Home Is Someone Else's Broken-Down Van

In the wealthiest county in the United States — Santa Clara County, California, home to Apple, Google, and eight of America's fifty most expensive zip codes — people are living in broken-down recreational vehicles parked on public streets. They are not homeless in the traditional sense. They have a roof over their heads. But that roof belongs to someone else, and they pay for it every month.

These landlords — called "vanlords" by housing advocates and city officials — purchase dilapidated, often unregistered, uninsured, and inoperable RVs, then rent them to people with nowhere else to go. The monthly fees vary, but reports from across California's Bay Area confirm that renters often pay hundreds of dollars per month — sometimes well over a thousand — for a vehicle that cannot legally be driven, often lacks working plumbing, and sits on a public street with no tenant protections whatsoever.

In Santa Clara County, the share of homeless individuals sleeping in vehicles has more than doubled since the pandemic, jumping from 18% in 2019 to 37% in 2025. Countywide, approximately 11,500 people are estimated to live in 6,800 RVs.

There are no leases. There are no inspections. There are no eviction processes — because legally, the renters have no rights to the vehicles they sleep in. City officials across California, including Los Angeles, San Jose, and San Francisco, have described the conditions in these RVs as hazardous: fire risks, biohazards, no sanitation, broken locks. One San Jose city councilmember described vanlords as individuals who "manipulate our unsheltered residents to rent these places in unsafe conditions."

The Economics of No Other Option

Why would anyone pay for this? Because in California — and increasingly across the United States — it is the only thing they can afford. California accounts for nearly one quarter of America's total homeless population despite having only 12% of its total residents. In the Bay Area, rents for a modest one-bedroom apartment routinely exceed $3,000 per month. Workers earning $25 per hour — which is above California's minimum wage — still cannot afford this.

Many of the people living in vanlord RVs are employed. Some are immigrants recently arrived in the country. Some are elderly. Some are families with children. They are not living on the street by choice; they are living in someone else's broken vehicle because it is one step above the street — and someone has found a way to profit from that one step.

An investigation by CNBC in February 2025 spoke directly with a vanlord and multiple tenants. One tenant reported paying several hundred dollars per month for a vehicle on a public street — an amount that, annualized, rivals what a decent apartment would cost in many parts of the United States.

The Policy Capture Problem: Who Protects the Renter?

Landlords in California must comply with detailed habitability requirements — working heat, plumbing, pest control, smoke detectors, and more. Vanlords comply with none of these requirements, because the vehicles are not legally classified as housing. This regulatory gap is not an accident; it is the result of what economists call "policy capture" — when the rules are written in ways that benefit those with money and power, while leaving vulnerable people without protection.

In 2023, the Los Angeles City Council voted 12-0 to study a crackdown on vanlords. L.A. County supervisors followed. San Jose proposed a ban on RV rentals to the homeless. Advocates for the unhoused pushed back — not to defend vanlords, but because they argued that removing even this deeply flawed housing option, without replacing it with anything better, would simply push people onto the bare pavement.

One resident at a San Jose RV encampment said it directly: "By restricting the sale, rental or even transfer of RVs, you're targeting one of the last remaining options for shelter that unhoused individuals can afford. If safety is the concern, then provide sanitation services in designated areas with proper oversight — not further displacement."

As of 2025, California has invested billions of dollars in homeless response with mixed results. The average time to move someone from an encampment to interim housing is three months. Many never make it. Meanwhile, the housing market continues to price people out, and vanlords continue to fill the gap — at whatever price desperation will bear.

 

🔍 Socratic Seminar Questions

1. Is a vanlord doing something wrong, or are they simply responding to a market need? Does your answer change if the RV is unsafe and uninhabitable?

2. Housing advocates say that banning vanlords without providing alternatives just moves people from one dangerous situation to another. Is this a valid argument against the ban? What would a better solution look like?

3. California has some of the highest minimum wages in the country and some of the highest rents. What does this tell us about what minimum wage laws can and cannot fix?

4. What is "policy capture," and can you find other examples of it beyond housing? (Think about other industries where regulations seem to protect corporations more than consumers.)

5. One councilmember said the city "does not want people living on the street, in an RV, or in a tent — but does not want to build enough housing to replace what is lost." If this is true, what is the city's actual goal?

6. Some of the people living in vanlord RVs are full-time workers. What does it mean for a society when working full time is no longer enough to afford a home?

💭 Food for Thought

• In 19th-century England, factory owners housed workers in company tenements and charged them rent — effectively taking back a portion of every paycheck. Historians call this the "truck system." Is vanlording a modern version of the same idea?

• Silicon Valley has produced more billionaires per square mile than almost anywhere on earth. It also has one of the worst housing crises in the developed world. Can both of these things be true at the same time without one causing the other?

• The government often describes homelessness as a personal failing — a result of addiction, mental illness, or bad decisions. The data from California suggests it is increasingly a result of economics: people simply cannot afford housing. Which explanation do the facts support? Does it matter which one politicians believe?

• Cities fine people for sleeping on sidewalks and ban RV rentals. They do not fine corporations for keeping wages low or fine landlords for leaving apartments empty as investments. What do these policy choices reveal about whose interests the government is protecting?

📚 Verify the Facts — Sources to Check

CNBC (February 20, 2026): "From 'vanlords' to safe parking sites: How RVs became Silicon Valley's housing safety net." cnbc.com

LAist / KPCC (October 2023): "LA County Supervisors Approve 'Vanlord' Resolution." laist.com

California City News (September 2023): "Los Angeles to Crack Down on So-Called 'Vanlords.'" californiacitynews.org

San José Spotlight (March 2025): "San Jose to crack down on RV rentals to homeless." sanjosespotlight.com

Greater Los Angeles Homeless Count (2023) — Published annual data on vehicle homelessness. lahomeless.org

 

PASSAGE 3  |  The Cycle of Collapse

Boom, Bust, Bailout, Repeat: How American Greed Has Repeatedly Crashed the Economy

A Nation That Does Not Learn Its Lessons

The United States economy has experienced three devastating bubble collapses in the last 25 years: the dot-com crash of 2000, the housing collapse of 2008, and the stock market crash that accompanied it. In each case, the pattern was identical: extraordinary profits for a small group of insiders, ordinary Americans left holding the wreckage, and the same forces that caused the crash being allowed to regroup and do it again.

This is not a conspiracy theory. It is documented history, laid out in detail by Nobel Prize-winning economists, the U.S. Senate, and the Financial Crisis Inquiry Commission — the official government body created to investigate the 2008 crash. Their conclusion: the financial crisis was "avoidable" and was caused by "widespread failures in financial regulation and supervision" and "dramatic failures of corporate governance and risk management." In plain language: people knew what was happening, made enormous amounts of money from it, and when it collapsed, ordinary Americans paid the price.

The Dot-Com Bubble: When Hype Became a Business Model

In the late 1990s, the invention of the internet created a genuine technological revolution. But the excitement around that revolution outpaced its reality. Investors poured money into internet companies — called "dot-coms" — many of which had no profits, no sustainable business models, and in some cases, no actual product beyond a website. Stock prices soared on pure speculation. Companies like Pets.com raised hundreds of millions of dollars in investor money before going bankrupt within a year of their initial public offering.

When the bubble burst in 2000, the Nasdaq stock index — where most tech stocks traded — fell approximately 78% from its peak. An estimated $6.2 trillion in household wealth was destroyed over two years. Small investors who had put their retirement savings into tech stocks lost everything. The executives and investment bankers who had hyped these stocks and collected massive fees? Most of them kept their money.

Adjusted for inflation, the dot-com crash led to losses of approximately $9 trillion in total wealth — more than the entire annual output of the U.S. economy at that time.

The Housing Bubble: The Crash That Shook the World

The dot-com crash led the Federal Reserve to slash interest rates, which made borrowing cheap. That cheap money flowed into the housing market. Between 1998 and 2006, the price of the average American home increased by 124%. Banks and mortgage lenders, sensing enormous profits, began issuing mortgages to people who could not realistically repay them — a practice known as "predatory lending." These loans were called "subprime mortgages."

What happened next is crucial to understand. The banks did not hold these risky loans. They bundled them together into complex financial products called "mortgage-backed securities" and sold them to investors around the world. Rating agencies — companies paid by the banks to assess the risk of these products — gave them top safety ratings, often with little real analysis. Why? Because the banks were paying them, and the banks wanted their products to look safe.

This is one of the purest examples of what economists call a "conflict of interest" — a situation where the person responsible for giving you an honest assessment is being paid by the person who benefits from a dishonest one.

In 2004, the Securities and Exchange Commission loosened rules that limited how much banks could borrow against their assets. Investment banks including Bear Stearns, Lehman Brothers, and Merrill Lynch began borrowing at ratios of 30-to-1 — meaning for every $1 they actually owned, they were betting $30 of borrowed money. When housing prices began to fall in 2006 and 2007, these institutions had nowhere to hide.

Lehman Brothers — once the fourth-largest investment bank in the United States — collapsed in September 2008, triggering the largest stock market crash since the Great Depression. The Dow Jones Industrial Average fell more than 50% from its peak to its lowest point in early 2009. Nearly 9 million Americans lost their jobs. Up to 10 million lost their homes. U.S. housing prices fell nearly 30% on average.

The government response: a $700 billion bailout of the financial sector, known as the Troubled Asset Relief Program (TARP). The banks were saved. The homeowners who had been given predatory loans were not. One analyst summarized it clearly: "Wall Street came out much better than Main Street."

Policy Capture: When the Regulators Work for the Regulated

A key factor in both collapses was what researchers call "regulatory capture" — the process by which the agencies created to oversee an industry are gradually taken over, ideologically or politically, by that industry's interests. In the years before 2008, the Federal Reserve, despite having the legal authority to regulate mortgage lending, declined to do so. The SEC loosened leverage rules at exactly the moment when leverage was becoming most dangerous. The rating agencies gave safe ratings to products they knew were risky.

How does this happen? Through lobbying. Through campaign contributions. Through the revolving door between government and industry, where regulators retire to the industries they once oversaw — and where industry executives take government positions that allow them to weaken the rules that once constrained them.

From 1998 to 2008, the financial industry spent more than $5 billion on lobbying and campaign contributions in Washington. The result: less regulation, fewer protections for consumers, and eventually, a global economic catastrophe whose costs were borne by the people least responsible for it.

What Comes Next? The AI Bubble

History does not repeat itself exactly, but it rhymes. As of 2025-2026, trillions of dollars are being invested in artificial intelligence companies, many of which have not yet demonstrated sustainable profits. Nvidia's stock at various points in 2024 commanded a price-to-earnings ratio — the measure of how much investors are paying for each dollar of actual profit — of well over 60, meaning investors are paying $60 for every $1 the company actually earns. By comparison, the broader stock market's historical average is closer to 15 to 20. The hype around AI is real and the technology is powerful. But the same things were said about the internet in 1999.

Whether the current AI investment surge becomes a bubble or a justified transformation remains to be seen. But the pattern — extraordinary profits for early insiders, valuations disconnected from current reality, and ordinary Americans' retirement funds and savings exposed to the risk — is familiar. Students studying this period of history should ask: who profits if the predictions come true? And who pays if they don't?

 

🔍 Socratic Seminar Questions

1. The dot-com bubble and the housing bubble both followed a similar pattern: excitement, overvaluation, collapse. Why do you think these patterns repeat, even when the previous crash is still in living memory?

2. Rating agencies gave toxic mortgage-backed securities top safety ratings because the banks were paying them. Can you think of other situations where the entity paying for an assessment has an incentive to receive a positive one? (Think: college rankings, restaurant reviews, product safety testing.)

3. In 2008, the government bailed out the banks but not the homeowners. Was this the right choice? What would have happened if the government had done the opposite — helped homeowners but let the banks fail?

4. What is a "conflict of interest"? Can you find three examples of conflicts of interest in the financial crisis story? Can you find one in your own school or community?

5. Some economists argue that bubbles are an inevitable part of capitalism — that human psychology naturally leads to cycles of boom and bust. Others argue that better regulation can prevent them. Which side does the evidence from this passage support?

6. Do you think the AI investment surge we are seeing today is a bubble, a justified revolution, or both? What evidence would you need to know for sure?

💭 Food for Thought

• The people who designed and sold the toxic mortgage products that caused the 2008 crash were mostly not prosecuted. Not a single senior executive from a major Wall Street bank went to prison. Does the absence of legal consequences change how you think about what happened — or about who is likely to do it again?

• The U.S. government printed and borrowed trillions of dollars to save the financial system in 2008. This money was ultimately paid for by taxpayers over years and decades. Who received the benefit? Who paid the cost? Is this a form of redistribution?

• Economists call the cycle of stability leading to risk-taking leading to collapse the "Minsky moment," after economist Hyman Minsky. His argument was simple: the longer things are stable, the more confident people become, and the more recklessly they behave — until they can't. If this is true, is the next crash not a matter of if, but when?

• The same financial firms that caused the 2008 crash — Goldman Sachs, JPMorgan, Citigroup — are today among the largest investors in AI. Does knowing this change how you think about the AI investment boom?

📚 Verify the Facts — Sources to Check

Financial Crisis Inquiry Commission (2011). "The Financial Crisis Inquiry Report." fcic.law.stanford.edu — The official U.S. government investigation. Freely available online.

Wikipedia: "2008 Financial Crisis" — Extensively sourced summary with original references. wikipedia.org/wiki/2008_financial_crisis

FiveThirtyEight (2014): "Why the Housing Bubble Tanked the Economy and the Tech Bubble Didn't." fivethirtyeight.com

Council on Foreign Relations: "The U.S. Financial Crisis" — Timeline of the crisis. cfr.org

Retro Report (2025): "The 2008 Financial Crisis Explained: Housing Bubble to Bailout." retroreport.org

 

PASSAGE 4  |  The Pattern

Greed Nation: Connecting the Dots

What Do a Diabetes Drug, a Broken-Down RV, and a Global Recession Have in Common?

At first glance, Ozempic, vanlords, and the 2008 financial crisis may seem like unrelated stories. But they share a common structure — a pattern that has appeared again and again in American economic history, and which, if students learn to recognize it, becomes visible in nearly every major public debate about money, power, and who benefits.

The pattern works like this:

Step 1: A company or industry identifies a human need — health, shelter, retirement security, communication.

Step 2: The company gains a dominant or monopoly position through patents, market power, or the elimination of competition.

Step 3: It uses that position to charge the maximum possible price — not the fair price, not the price that reflects actual cost, but the highest price it can extract before people either give up or die.

Step 4: It deploys money — through lobbying, campaign contributions, and the revolving door between industry and government — to prevent regulation that might limit its power.

Step 5: When the extraction causes a crisis — a health disaster, a housing emergency, an economic collapse — the cost is transferred to the public, while the private profits remain private.

Novo Nordisk extracts profit from diabetics. Vanlords extract profit from the unhoused. Wall Street extracted profit from homeowners and pension funds — and when the entire system collapsed, it extracted $700 billion more from taxpayers. The mechanism is the same. Only the industry changes.

The Question of Complicity

One of the hardest parts of studying this pattern is confronting the question of complicity. These systems do not run themselves. They require lawyers who draft the patent filings. They require accountants who structure the profit margins. They require politicians who accept the campaign contributions and decline to act. They require regulators who choose not to regulate. They require economists who write papers justifying the status quo. They require journalists who don't run the story, and voters who don't demand accountability.

None of these people think of themselves as villains. Most of them are doing their jobs, following the incentives they have been given, operating within systems they did not create. This does not make the outcomes less real. It raises a harder question: How do you change a system when almost everyone inside it is benefiting from it in some way, or simply has too much to lose by challenging it?

What Students Can Do

This series is not designed to make you feel hopeless. It is designed to make you see clearly. The first step in changing any system is understanding how it actually works — not how it claims to work, not the version taught in a textbook designed to make existing institutions look neutral and rational, but the actual mechanics of power and profit.

You will graduate into an economy shaped by these forces. You will pay for insurance, take medications, apply for housing, invest in retirement, and vote for people who make decisions about all of these things. The more clearly you understand what is happening and why, the less easily you can be manipulated by the language that is used to justify it — words like "innovation," "market efficiency," "property rights," and "free enterprise." These words are not wrong. But they are often used as shields behind which very specific interests are protected.

The question is not whether greed exists. Of course it does. The question is whether we build systems that harness it productively, limit its worst excesses, and distribute its benefits broadly — or whether we build systems that allow it to feed on human need without limit or accountability.

That question is not settled. It is being answered, right now, by the choices of the generation you belong to.

 

🔍 Socratic Seminar Questions

1. In your own words, describe the five-step pattern identified in this passage. Can you find a current news story that fits this pattern?

2. The passage says that the people who sustain these systems often don't think of themselves as villains. Do you find this convincing? Does intent matter when the outcome causes harm?

3. The passage lists several groups as "complicit" in maintaining systems of exploitation: lawyers, accountants, politicians, regulators, economists, journalists, voters. Is it fair to include all of these groups? Are any of them more responsible than others?

4. What is the difference between "free enterprise" as a concept and the actual market conditions described in these passages? Can you have free enterprise without competition?

5. If you were designing an economic system from scratch, what rules would you put in place to prevent the patterns described in this series? What problems might your rules create?

6. The series ends with the claim that your generation will answer the question of what kind of economy we build. Do you believe this? What would it actually take to change the systems described here?

💭 Food for Thought

• Economist Milton Friedman famously argued that a corporation's only responsibility is to maximize profits for its shareholders. Economist Joseph Stiglitz and others have argued that this idea, taken to its logical conclusion, produces exactly the outcomes described in this series. Who do you think is right — and what evidence from these passages supports your view?

• The United States spends more on healthcare per person than any other wealthy nation, and yet has worse health outcomes on most major measures. It has some of the highest housing costs among developed nations. It has the largest wealth gap of any G7 country. Are these facts connected to the patterns described in this series, or are they coincidences?

• History shows that major reforms — child labor laws, the 40-hour workweek, Social Security, civil rights legislation — happened not because powerful interests willingly gave up power, but because ordinary people organized, demanded change, and sometimes paid a price for it. What would it take for a similar shift to happen now? What stands in the way?

• We use the word 'greed' as though it is a character flaw — something that bad people have. But economist Adam Smith, who is considered the father of modern capitalism, argued that self-interest, properly channeled, produces good outcomes for society. At what point does self-interest become greed? And who decides?

 

SECTION 3: What Is a Socratic Seminar?

An Overview for Educators and Students, Based on Socratic Seminars International

The Idea Behind the Method

A Socratic seminar is a structured intellectual discussion in which a group of participants — students, adults, or anyone seeking understanding — explore a shared text through carefully prepared open-ended questions. The method draws its name from the Greek philosopher Socrates, who taught not by lecturing but by questioning: asking his students to examine their assumptions, define their terms, and follow an argument to its logical conclusion, even when that conclusion was uncomfortable.

Socratic Seminars International, founded by educator Oscar Graybill, formalized and expanded the method for use in modern classrooms. Their core definition: a Socratic seminar is "a collaborative, intellectual dialogue facilitated with open-ended questions about a text." The emphasis is on dialogue — not debate. In a debate, participants defend fixed positions and try to win. In a Socratic seminar, participants explore multiple perspectives and try to understand.

The goal of a Socratic seminar is not agreement. It is not the defeat of one argument by another. It is a deeper, more nuanced understanding of difficult ideas — the kind of understanding that only emerges when people are honest about what they don't know and genuinely curious about what they might be missing.

How It Differs from a Typical Classroom Discussion

TYPICAL CLASS DISCUSSION • Teacher asks questions; students answer • Conversation flows through the teacher • Right answers are the goal • Students respond to earn participation credit • Teacher evaluates accuracy of responses • Passive listening between turns

SOCRATIC SEMINAR • Students ask and answer each other's questions • Conversation flows directly between students • Better questions are the goal • Students respond because they are genuinely curious • Participants assess quality of the dialogue collectively • Active listening is a core skill being practiced

Three Types of Questions

Socratic Seminars International identifies three levels of questions, each serving a distinct purpose in deepening understanding. These correspond roughly to Benjamin Bloom's taxonomy of learning, and a well-run seminar moves through all three:

1	Opening / Grounding Questions These connect the text to the student's own experience. They are designed to give everyone an accessible entry point and establish that the conversation is rooted in real life. Example: 'Have you or has someone you know ever been unable to afford a medication? How did that feel?' These questions open the seminar and ensure no one starts from silence.
2	Core Analytical Questions These push students into the text itself — asking them to analyze claims, evaluate evidence, identify assumptions, and examine the author's reasoning. Example: 'The passage states that Novo Nordisk's drug can be made for $4.73 per month but costs nearly $1,000. What assumption must be true for this pricing to be considered legitimate?' These are the heart of the seminar.
3	Closing / Synthesis Questions These expand the conversation beyond the specific text to larger principles, patterns, or implications. Example: 'If the pattern described in this series repeats across tobacco, lead, PFAS, and pesticides, what does this tell us about the reliability of corporate self-regulation?' Closing questions send students out of the room still thinking.

The Recipe for Success (Socratic Seminars International)

Oscar Graybill and Socratic Seminars International describe what they call a "Recipe for Success" for facilitating effective seminars. The key elements:

ELEMENT — WHAT MAKES A SOCRATIC SEMINAR WORK

A rich, challenging text — not too long to read closely, not so simple it yields no complexity

Preparation — students must have read the text before arriving; annotation is strongly encouraged

Open-ended questions — questions whose value lies in their exploration, not their answer

A circle arrangement — everyone must be able to see everyone else; hierarchy is physically reduced

Ground rules established in advance — respect, evidence, no personal attacks, listening before responding

A facilitator, not a teacher — the leader asks opening questions and steps back; students drive

Reflection — after every seminar, participants evaluate the quality of the conversation, not just its content

No bad seminar if you reflect — even a seminar that struggles teaches the group something about dialogue

The Fishbowl Format (Recommended for These Readings)

For classes of 20 or more students, Socratic Seminars International and most classroom practitioners recommend the fishbowl format:

1	Set Up Two Circles Arrange chairs in an inner circle (speakers) and an outer circle (observers). The inner circle holds the active discussion. The outer circle observes, takes notes, tracks participation, and prepares to swap in.
2	Inner Circle Discusses The inner circle (typically 8–12 students) discusses the text using prepared questions. The teacher/facilitator asks one opening question and then steps back. Students call on each other, build on each other's ideas, and refer back to the text.
3	Outer Circle Observes Outer circle students use a structured observation form: tracking which students speak, noting strong arguments, recording unanswered questions, and preparing their own contributions for when they rotate in.
4	Rotate and Continue After 15–20 minutes, circles switch. New inner circle students bring fresh questions informed by what they observed. Repeat.
5	Full Group Debrief Close with 10 minutes of whole-group reflection: What was the strongest argument made? What question still hasn't been answered? What would you want to say now that you didn't say during the seminar?

Ground Rules for Students

Post these or distribute them before every seminar. Adapted from Socratic Seminars International and NCTE guidelines:

SOCRATIC SEMINAR GROUND RULES

Refer to the text — support every major claim with evidence from the reading

Listen before responding — do not begin formulating your reply while someone else is still speaking

Build, don't just react — acknowledge what the previous speaker said before adding your own idea

Ask questions more than you make statements — curiosity is more valuable than certainty here

Invite quieter voices — if someone has not spoken, invite their perspective

No put-downs — challenge ideas, never people; disagree with reasoning, not with the person

Be willing to change your mind — the best seminars end with people thinking differently than when they began

Stay in the text — personal anecdotes are welcome as illustrations, not as substitutes for evidence

Assessing Socratic Seminars

Socratic Seminars International is clear: the most important measure of success is not whether students reached correct conclusions, but whether the dialogue itself was substantive, text-grounded, and collegial. Reflection is described as the key to improvement.

Suggested assessment approaches for these reading series:

ASSESSMENT APPROACH — OPTIONS FOR EDUCATORS

Self-assessment: Students rate their own participation on dimensions of listening, evidence use, and question quality

Peer assessment (outer circle): Observers track speakers and note strongest contributions using a structured rubric

Pre/post writing: Students write a 200-word response to the opening question before the seminar; revise it after

Exit ticket: One question they still have; one argument they found most persuasive; one idea they changed their mind about

Quality of questions: Evaluate student-written questions on whether they are open-ended, text-grounded, and generative

Process over content: Grade on demonstrated listening (did students build on each other?) rather than correctness of views

 

SECTION 4: How to Read Closely — The Mortimer J. Adler Method

Based on How to Read a Book by Mortimer J. Adler and Charles Van Doren (1940; revised 1972, Simon & Schuster)

Why Adler Matters for These Readings

Mortimer J. Adler was an American philosopher and educator who spent his career arguing that reading is a skill most people are never fully taught — and that genuine understanding requires far more than decoding words on a page. His 1940 book How to Read a Book, revised in 1972 with Charles Van Doren, remains one of the most practical and influential guides to active, critical reading ever written.

The passages in Greed Nation and Poisoned for Profit are designed to be read, not skimmed. They contain layered arguments, verifiable claims, and deliberate tensions. Applying Adler's method to these passages will help students get far more from each reading — and will prepare them to participate in Socratic seminars with the kind of specific, text-grounded understanding that makes those seminars valuable.

Adler's core argument: Most people read to be entertained or to collect information. Very few people read to understand. Understanding — the kind that changes how you think — requires active effort, good questions, and a willingness to argue back at the page.

The Four Levels of Reading

Adler identifies four levels of reading, each building on the previous. Think of them as reading to decode, reading to survey, reading to understand, and reading to master.

Level	Name	Core Question	What You Do
1	Elementary Reading	What does this sentence say?	Decode the words. Understand basic grammar and vocabulary. Identify unfamiliar terms.
2	Inspectional Reading	What is this text about? What is its structure?	Skim quickly. Read titles, headings, callout boxes. Read first and last paragraphs. Form a mental map before reading carefully.
3	Analytical Reading	What does this text mean? Is it true?	Read deeply and completely. Identify the author's argument, key claims, and evidence. Evaluate logic and evidence quality. Argue back at the text.
4	Syntopical Reading	How does this compare to other texts on the same subject?	Read multiple texts on the same topic. Compare claims, identify agreements and conflicts, develop your own position through synthesis.

For these reading series, students should be working at Level 3 (Analytical Reading) for each passage, and by the final synthesis passage of each series, they should be working at Level 4 — comparing arguments across all the passages they have read.

Adler's Four Core Questions for Every Text

Adler argues that every serious reader must ask four questions of every text they read analytically. These are not optional. They are the work:

1	What is this text about, as a whole? Not a summary of every paragraph — a single sentence or two capturing the central argument or main idea. If you cannot state it in one sentence, you do not yet understand the text well enough to discuss it. Before your Socratic seminar, write this sentence in the margin or at the top of your notes.
2	What is being said, in detail? What are the key claims and how are they supported? Identify the three to five most important specific claims in the text. For each one, ask: What is the evidence? Is the evidence sufficient? Could the evidence be interpreted differently? Mark these passages. They are your ammunition for the seminar.
3	Is this text true, in whole or in part? This is the hardest question and the one most readers skip. Do not simply accept what you have read. Ask: Does this argument follow logically? Are the sources reliable? Does the evidence actually support the conclusion? Are there facts that seem designed to provoke rather than to inform? Even a passage you largely agree with deserves this scrutiny.
4	What of it? Why does this matter? What are the implications of this text if it is true? What would have to change — in policy, in behavior, in how you think about the world — if everything in this passage is accurate? This is the question the Socratic seminar's 'Food for Thought' sections are designed to push you toward.

Adler's Rules for Active Reading

Adler is explicit: a book (or passage) you have not written in is one you have not really read. He recommends treating annotation as a conversation with the author — a way of arguing back, asking questions, and making the ideas your own. For these reading series, students should annotate every passage before the seminar.

ADLER'S ANNOTATION SYSTEM — ADAPT THESE FOR EACH PASSAGE

Underline main ideas and key arguments — one main idea per paragraph maximum

Circle unfamiliar or significant words and look them up before the seminar

Write a question mark (?) next to anything you don't understand or don't believe

Write an exclamation mark (!) next to anything that surprises or disturbs you

Star (*) the one sentence per passage that you think is most important

Write a brief summary (2–3 words) in the margin next to each major section

At the end: write your answer to 'What is this passage about, as a whole?' in one sentence

Write your answer to 'Is this true?' — note at least one point of agreement and one point of doubt

Applying Adler to These Specific Passages

For Greed Nation and Poisoned for Profit, here is how to apply Adler's method before your Socratic seminar:

1	Before You Read: Inspectional Reading (5 minutes) Read the passage title, all section headings, the stat boxes and callout quotes, and the Socratic seminar questions at the end — before reading the body text. This gives your brain a framework. You will understand the detailed argument far better if you know where it is going.
2	First Read: Analytical Reading — Don't Stop (15–20 minutes) Read the entire passage through without stopping to look things up. Mark confusing passages with a question mark. Underline sentences that feel important. Keep moving. Adler is emphatic: the first read should be complete, even if parts are unclear.
3	Second Read: Active Annotation (10–15 minutes) Re-read more slowly, applying the annotation system above. For each major section, write a two-word summary in the margin. Circle the one claim per section you most want to question in the seminar. Write your question in the margin.
4	Verify the Sources (10 minutes, shared with class) Each passage includes a 'Verify the Facts' source box. Look up at least one source before the seminar. Does the source say what the passage claims it says? Is there context the passage left out? Finding a discrepancy — or confirming accuracy — gives you powerful material for discussion.
5	Write Your Opening Contribution (5 minutes) Before the seminar, write one sentence you intend to say in the first five minutes — either a question you want to raise, a claim you want to challenge, or an observation you want to test against the group. Having something prepared prevents the paralysis of the blank page and ensures you enter the conversation with purpose.

Syntopical Reading Across the Series

By the time students reach the synthesis passages at the end of each series, Adler would say they are ready for Level 4: syntopical reading — reading multiple texts on the same subject in order to develop their own position through comparison and synthesis.

After completing both series, students should be asked:

Across all nine passages (four in Greed Nation, five in Poisoned for Profit), what is the single most important pattern you see? Where do the series agree? Where do they diverge? If everything in both series is accurate, what is the most urgent implication for your generation — and what is one concrete thing that would need to change for that implication to be addressed?

This is the question Adler's method is designed to make answerable — not because there is one right answer, but because close reading of multiple sources prepares the mind to form an honest, well-reasoned position of its own. That is, as Adler would say, what it means to become a truly competent reader.

 

QUICK REFERENCE: Combining Adler + Socratic Seminar for These Readings

Print or post this for students to use as a preparation checklist.

THE COMPLETE PREPARATION CHECKLIST (STUDENT VERSION)

BEFORE YOU READ: Skim titles, headings, callout boxes, and the Socratic questions at the end

FIRST READ: Read all the way through. Mark confusing parts with '?'. Don't stop.

SECOND READ: Annotate. Underline main ideas. Circle key words. Star the most important sentence.

ANSWER ADLER'S QUESTION 1: Write one sentence: 'This passage is about ________________.'

ANSWER ADLER'S QUESTION 2: List 3 specific claims and their evidence. Do you believe the evidence?

ANSWER ADLER'S QUESTION 3: Write one thing you believe is true and one thing you want to challenge.

ANSWER ADLER'S QUESTION 4: If all of this is accurate, what should change? Why does it matter?

VERIFY A SOURCE: Look up at least one source from the 'Verify the Facts' box. Does it check out?

WRITE YOUR OPENING CONTRIBUTION: One sentence or question you will raise in the first 5 minutes.

DURING THE SEMINAR: Listen before speaking. Build on what others say. Refer back to the text.

AFTER THE SEMINAR: Write your exit ticket — one question you still have; one idea you revised.

References and Sources for This Package

Adler, Mortimer J. and Van Doren, Charles. How to Read a Book: The Classic Guide to Intelligent Reading. Revised edition. Simon & Schuster, 1972.

Socratic Seminars International. socraticseminars.com — Graybill's "Recipe for Success" and ground rules.

Israel, Elfie. "Examining Multiple Perspectives through Socratic Seminars." ReadWriteThink, readwritethink.org.

National Council of Teachers of English. "Crafting and Conducting a Successful Socratic Seminar." NCTE.org, 2017.

Facing History & Ourselves. "Socratic Seminar Teaching Strategy." facinghistory.org.

Spencer Education. "Designing Socratic Seminars to Ensure That All Students Can Participate." spencereducation.com, 2026.

This series was designed for use in Socratic Seminar discussions, where no one person holds all the answers and the quality of questions matters more

BONUS SEMINARS: 

THE CHAOS CODE

A Socratic Seminar Reading Series for High School Students

How Social Media, Big Tech, and AI Are Engineering Fear, Addiction, and Profit at the Cost of Your Mind

 

You did not choose to be addicted to your phone. The addiction was built for you — engineered by teams of behavioral scientists, neuroscientists, and machine learning systems whose sole job was to figure out what keeps you scrolling one second longer. That extra second is worth money. Billions of seconds, from billions of users, add up to billions of dollars in advertising revenue. And the currency those seconds are extracted with is not entertainment or connection. It is fear. It is anger. It is the particular dread that something important is happening right now, and you are missing it.

This reading series examines four interlocking forces that are reshaping how young people — and everyone else — experience reality: the algorithms engineered to keep you in a state of emotional agitation, the collapse of online information into AI-generated garbage designed to generate clicks rather than inform, the conversion of every digital tool into a subscription that extracts money indefinitely, and the concentrated power of a handful of billionaires who built these systems and who are now using them in ways that go far beyond entertainment.

These are not opinions. They are documented engineering decisions, published research findings, internal documents released through litigation and whistleblower testimony, and Senate hearing records. The sources are listed. Check them. These facts belong to you.

 

PASSAGE 1  |  The Algorithm of Outrage

Engineering Anger: How Social Media Platforms Are Designed to Keep You Afraid and Scrolling

The Casino in Your Pocket

In 1953, psychologist B.F. Skinner discovered something that would later become the foundation of the most profitable business model in human history. He found that laboratory rats pressed a lever most compulsively — not when they received a food reward every time, but when rewards came randomly and unpredictably. He called this "intermittent reinforcement." It is the same mechanism that makes slot machines the most profitable gambling device ever invented. It is also the core design principle behind every major social media notification system on earth.

The engineers who built Facebook's "Like" button understood this. Former Facebook President Sean Parker — one of the company's earliest executives — has stated publicly that the platform was built to exploit "a vulnerability in human psychology." He said the design question they asked themselves was: "How do we consume as much of your time and conscious attention as possible?" Their answer: give people "a little dopamine hit every once in a while" by showing them likes and comments on their posts, which creates "a social validation feedback loop" — exactly like a slot machine. Parker made these comments in 2017, after leaving Facebook, noting that "God only knows what it's doing to our children's brains."

Neta Alexander, an assistant professor at Yale who co-teaches a seminar called "Media Anxieties," describes social media platforms as "designed to be addictive by using intermittent rewards and trying to invoke negative emotional responses such as rage, anxiety and jealousy, which are known to prolong our engagement and deepen our attachment to our devices." This is not a side effect of the design. It is the goal.

Why Fear and Anger Are the Algorithm's Favorite Emotions

Not all emotions are equal in the attention economy. Research published in Psychological Science in 2024 found that constant exposure to emotionally charged social media content heightens stress, anxiety, and feelings of paranoia — and that content eliciting extreme emotional responses, particularly fear and anger, keeps users scrolling the longest. Platforms discovered this not through academic research but through their own internal data: angry, anxious users click more, comment more, share more, and return more frequently than calm, satisfied users.

A 2021 internal Facebook study — leaked through whistleblower Frances Haugen and reported by the Wall Street Journal — found that the company knew its algorithm was amplifying "anger and outrage" and that this was a direct result of a 2018 change to its ranking system designed to boost "meaningful social interaction." When the company studied the change, it found that it was increasing exposure to divisive, inflammatory content. An internal researcher wrote that the change "may be one of the most powerful drivers of misinformation." Facebook was shown this research. It proceeded anyway.

145 min

The average time per day that humans worldwide now spend on social media — the equivalent of more than 37 full days per year spent in an engagement-optimization machine.

The neuroscience underneath this is increasingly well-documented. Research published in PMC (the National Institutes of Health's open-access journal archive) in 2025 confirmed that frequent social media engagement alters dopamine pathways in ways "analogous to substance addiction." Brain imaging studies show changes in the prefrontal cortex — the part of the brain responsible for impulse control and rational decision-making — and in the amygdala, which processes threat and emotional response. Heavy social media users show increased emotional sensitivity and compromised decision-making. These are not metaphors. They are measurable changes in brain structure and function produced by an app on a phone.

The Infinite Scroll: Designed to Have No End

The infinite scroll — the feature that causes your social media feed to load new content automatically, with no natural stopping point — was invented by Aza Raskin, a designer at a tech company in the late 2000s. He has since publicly expressed regret about the invention. In a 2018 interview, Raskin estimated that infinite scroll causes users to spend an additional 200,000 hours on social media every day — across all users — compared to a design that had natural stopping points. He has said: "It's as if they took behavioral cocaine and sprinkled it on the screen." He now works on technology ethics and has said the industry has caused "a race to the bottom of the brain stem."

Infinite scroll is not alone. "Autoplay" — YouTube's feature that automatically plays the next video — was shown in internal Facebook research to be a significant driver of radicalization, as the algorithm progressively serves more extreme content in search of higher engagement. The notification system — the red badge on your app icon, the vibration of your phone, the sound that signals someone has responded to you — is designed to create what behavioral scientists call "checking behavior": compulsive, semi-involuntary rechecking of the app at intervals throughout the day. In a Yale class experiment, one in ten students could not stay off social media for 24 hours even after voluntarily committing to do so. One student touched the Instagram icon involuntarily — as a reflexive motor habit — before catching themselves.

According to Pew Research Center, 59 percent of U.S. teenagers report feeling pressure to look good or appear successful on social media. A two-week digital detox — reducing social media to 30 minutes per day — significantly reduced anxiety, depression, loneliness, and FOMO scores in clinical studies. The platforms know this. The limitation is not a feature they offer.

Who Is Responsible?

The companies that built these systems — Meta (Facebook, Instagram, WhatsApp), Alphabet (YouTube), TikTok, Snapchat, and X (formerly Twitter) — have, in every public-facing statement, described their platforms as tools for connection, community, and free expression. In testimony before Congress, their executives have said they take the mental health of young users seriously. In internal documents, emails, and studies made public through litigation and whistleblower testimony, a different story appears: companies knowingly used design mechanisms that exploited psychological vulnerabilities, observed the mental health harms, and continued.

In January 2024, the CEOs of Meta, TikTok, X, Discord, and Snapchat were called before the United States Senate Judiciary Committee. Parents of children who had died — from suicide linked to Instagram use, from online extortion, from exposure to harmful content — sat behind the executives. Senator Lindsey Graham said directly to Mark Zuckerberg: "You have blood on your hands." Zuckerberg turned to the gallery of grieving parents and said he was sorry. No laws were passed. No executives faced personal legal liability. The platforms continued operating under the same algorithmic structures that produced the outcomes described in those parents' testimony.

 

🔍  Socratic Seminar Questions

1.  Sean Parker said Facebook was designed to exploit "a vulnerability in human psychology." He helped build it anyway. At what point does building something you know is harmful become a moral failure — or a legal one?

2.  The Facebook internal study showed the algorithm amplified misinformation and divisiveness. The company chose not to change it. Is this different from a car company knowing its brakes are defective and choosing not to issue a recall? Why or why not?

3.  Aza Raskin invented the infinite scroll and now campaigns against it. He knew it was psychologically manipulative when he built it. How much responsibility does an individual engineer bear for the large-scale consequences of a design decision they were paid to make?

4.  Research shows that limiting social media to 30 minutes per day significantly improves mental health outcomes. Platforms know this and do not offer it as a default. Why not? What does the answer tell you about whose interests the platform serves?

5.  The Senate hearing with tech CEOs produced no legislation and no personal liability. What does this tell us about whether our existing political and legal systems are equipped to address harms caused by technology companies?

6.  If you discovered that a drug company had internal research showing its product caused addiction and depression, and chose to continue selling it, you would probably describe that as criminal. Is the social media situation meaningfully different? If so, how?

💭  Food for Thought

•  The behavioral mechanisms used by social media platforms — intermittent reinforcement, variable reward, FOMO engineering — were first developed to understand and treat addiction. The same science that helps therapists treat gambling disorder was later used to maximize engagement on Facebook. What does it mean that the tools of healing and the tools of exploitation are identical?

•  B.F. Skinner's rats did not know they were in a Skinner box. Do you? The argument for platform regulation is partly that users cannot give truly informed consent to systems they do not understand. The argument against is that adults have the right to make their own choices. Which side does the evidence support — and does your answer change when the users are 13 years old?

•  Social media platforms are free to use. This is their core value proposition. But the actual cost — paid in attention, data, and mental health — is not disclosed in any terms of service. If the full cost of using Instagram were printed on the login screen the way cigarette warning labels are printed on packs, what would it say? Write it.

•  Every major social media company says its algorithm shows you what you want to see. But internal research shows the algorithm shows you what keeps you engaged the longest — which is not the same thing. If you could design the algorithm, what would you optimize for instead of engagement? What would be lost? What would be gained?

📚  Verify the Facts — Sources to Check

Parker, Sean (November 2017). Public remarks at Axios event: Facebook was built to exploit 'a vulnerability in human psychology.' Reported by The Guardian: theguardian.com

De et al. (2025). 'Social Media Algorithms and Teen Addiction: Neurophysiological Impact.' PMC / National Institutes of Health. pmc.ncbi.nlm.nih.gov/articles/PMC11804976/

Wall Street Journal (September 2021). 'The Facebook Files' — Frances Haugen whistleblower documents. wsj.com

Pew Research Center (2023). Teen social media and technology survey — 59% of teens feel pressure to look good online. pewresearch.org

Alexander, Neta. Yale Daily News (November 2024): 'Algorithmic Manipulation: How Social Media Platforms Exploit Student Vulnerabilities.' yaledailynews.com

Raskin, Aza (2018). Interview on infinite scroll design regrets. Reported by the BBC and multiple outlets. Raskin now leads Center for Humane Technology.

 

PASSAGE 2  |  The Slop Economy

AI Slop, Misinformation Farms, and the Death of Truth Online

The 2025 Word of the Year: Slop

In January 2025, the American Dialect Society selected "AI slop" as its Word of the Year. The choice was both a linguistic recognition and a cultural verdict. AI slop is the term for content generated by artificial intelligence — images, videos, articles, posts — that is produced in bulk for the sole purpose of generating engagement and advertising revenue, with no regard for accuracy, quality, or the effect on the people who consume it. It is not a minor or fringe phenomenon. It is currently one of the dominant forces shaping what people see, believe, and share online.

The mechanism is straightforward and, once understood, impossible to unsee. Social media platforms — Facebook, TikTok, Instagram, YouTube, and X — pay creators through engagement-linked advertising programs. Content that gets more clicks, more shares, more comments, and more time-on-screen generates more revenue. AI tools now make it possible to produce this content at industrial scale, at near-zero cost. The result: content farms staffed by operators in countries around the world produce hundreds of thousands of AI-generated posts per day, optimized algorithmically to trigger the exact emotional responses — outrage, grief, sentimentality, shock — that the engagement algorithms reward.

15B	The number of 'higher-risk' scam ads that Meta's own internal documents estimated its users were exposed to per day in 2024, according to reporting by Reuters. Advertising for scams and banned goods was projected to bring in 10% of Meta's total annual ad revenue that year.

The slop economy has a human face. A Facebook account that regularly posts AI-generated videos of an elderly man describing incontinent misadventures reportedly generates its operator upward of $5,000 per month through Meta's creator monetization program. "Shrimp Jesus" — an AI-generated image of a half-human, half-crustacean figure that went viral on Facebook in 2024 — became a symbol of the crisis: bizarre, meaningless, human-free content that the algorithm amplifies because it generates reactions. Facebook's feed, as of 2024-25, is estimated to be significantly composed of AI-generated or AI-assisted content, alongside a significant volume of outright scams. The company earned revenue from all of it.

The term 'enshittification' was coined by author and technologist Cory Doctorow to describe the lifecycle of digital platforms: they begin by being genuinely good for users to build a large audience; then they begin exploiting users to serve advertisers; then they degrade the product further to extract maximum profit before users have nowhere better to go. The word was added to major dictionaries in 2024. It describes exactly what has happened to Facebook, Google Search, and Amazon.

When AI Slop Becomes Dangerous

The consequences of an information environment flooded with AI-generated content range from embarrassing to genuinely deadly. In 2024, in the aftermath of Hurricane Helene, Republican influencer Laura Loomer shared an AI-generated image of a young girl clutching a puppy in floodwaters as supposed evidence of government failure. The image was fake. The activist who shared it acknowledged it was not real. It was shared hundreds of thousands of times. On a smaller but multiplied scale, this dynamic — emotionally compelling fake content amplified by algorithms and shared by real people before being fact-checked — is now a routine feature of every major breaking news event.

The consequences are not limited to politics. During the COVID-19 pandemic, AI-assisted misinformation about vaccines spread on social media platforms faster than official public health guidance. A study published in The Lancet found that social media misinformation was directly associated with lower vaccination rates. On mental health: algorithms that track users who engage with content about depression or self-harm have been shown to progressively serve those users more of the same content, deepening isolation rather than offering support. Instagram's own internal research, made public through Frances Haugen's 2021 disclosure, found that the platform made body image issues worse for one in three teenage girls — and that the company chose not to act on this finding.

1 in 3

Teen girls whose body image problems were made worse by Instagram use, according to Facebook's own internal research — research that company chose not to act on, according to documents disclosed by whistleblower Frances Haugen in 2021.

The Political Weaponization of Slop

AI slop has moved from content farms into the political mainstream. In 2025, Wired magazine described Donald Trump as "the first AI slop president" — documenting his use of AI-generated images on official government social media accounts depicting himself as a pope, a fighter pilot, and a muscular action hero wielding a lightsaber. In August 2024, Trump shared AI-generated images on Truth Social showing Taylor Swift fans in "Swifties for Trump" T-shirts and appearing to endorse his campaign. Swift had not endorsed Trump. The images were fabricated. They were shared by Trump's social media accounts to millions of followers.

The initial version of the Make Our Children Healthy Again Assessment — the health report issued by Robert F. Kennedy Jr.'s commission in 2025 — reportedly cited nonexistent references generated by AI. The AI had fabricated sources that did not exist. Those fabricated citations were in a document released by a cabinet-level U.S. government commission. More than half of the entries in Grok's AI-powered encyclopedia (Grokipedia, created by Elon Musk's xAI) were found to be at least partly copied from Wikipedia, with entries on controversial topics significantly rewritten to reflect a specific narrative — without disclosing to readers that the rewriting had taken place.

This is the endpoint of the slop economy: when the tools designed to generate engagement-optimized fake content are deployed not by anonymous content farms but by heads of state and cabinet officials, the informational infrastructure that democracy requires to function is directly threatened. A citizenry that cannot distinguish verified fact from AI-generated fabrication cannot make informed political decisions. That is not a side effect. For some of the people deploying these tools, it appears to be the goal.

 

🔍  Socratic Seminar Questions

1.  Meta earns revenue from scam ads and AI slop content alongside legitimate advertising. Is this meaningfully different from a newspaper that prints a fraudulent advertisement? Who bears the moral and legal responsibility for the harm caused?

2.  The platform algorithm that amplifies AI slop is the same algorithm that amplifies misinformation about vaccines, disasters, and elections. Is this a technology problem — something that better AI could fix — or a business model problem that better AI would only make worse?

3.  Instagram's internal research showed its platform worsened body image problems for one in three teenage girls. The company did not act on this finding. Using what you know about corporate decision-making from the previous reading series, why might a company choose not to act on evidence that its product is causing harm?

4.  Donald Trump shared AI-generated images of Taylor Swift endorsing his candidacy. His cabinet's health report cited sources that did not exist, apparently generated by AI. What should the legal consequences of this be — if any? What makes it difficult to hold public officials legally accountable for AI-generated misinformation?

5.  The word 'enshittification' describes platforms that start good, then progressively degrade their product to extract maximum profit. Can you think of other products or services — beyond digital platforms — that follow this same lifecycle? What does it suggest about what markets optimize for?

6.  You personally consume information online every day. What specific practices could you use, starting tomorrow, to identify whether content you are seeing is AI-generated, algorithmically amplified misinformation, or genuine? What is making that harder, and what is making it easier?

💭  Food for Thought

•  In the 1930s, Nazi Germany used film, radio, and posters to flood the information environment with a specific narrative — not primarily by making people believe lies, but by making people so confused and overwhelmed by competing claims that they stopped believing anything was true. Historian Hannah Arendt called this the goal of totalitarian propaganda: not the triumph of the lie, but the destruction of the category of truth itself. Does the current AI slop environment serve a similar function, intentionally or not?

•  Facebook earns approximately 10% of its total annual ad revenue from scam and fraudulent advertising, according to its own internal estimates. YouTube pays creators whose content generates engagement, regardless of whether that content is real or fabricated. The business model, in other words, is not accidental to the slop problem — it is its cause. If you wanted to fix the slop problem, would you fix the technology, or would you change the business model? Why?

•  A researcher who studies algorithmic radicalization on YouTube described the process as follows: the algorithm doesn't start showing you extremist content. It starts showing you content that's slightly more engaging than what you just watched. Over hundreds of sessions, 'slightly more engaging' leads to 'dramatically more extreme.' Can you think of any personal experience — with media, with friend groups, with any gradual change — that follows the same logic?

•  The American Dialect Society's 2025 Word of the Year is 'AI slop.' Every year, the words a culture chooses to name — or refuses to name — reveal something about what that culture is processing and what it is afraid of. What do you think the 2026 Word of the Year might be? What does your answer reveal about what you think is coming next?

📚  Verify the Facts — Sources to Check

American Dialect Society (January 2025). 'AI Slop' selected as 2025 Word of the Year. americandialect.org

Reuters (2024). Meta exposed users to approximately 15 billion higher-risk scam ads per day, per internal company documents. reuters.com

Haugen, Frances (October 2021). U.S. Senate testimony and document disclosures — Facebook internal research on teen mental health. Available via Senate Commerce Committee records.

Wikipedia: 'AI Slop' — extensively sourced entry covering political uses, content farms, and platform responses. en.wikipedia.org/wiki/AI_slop

Doctorow, Cory. Enshittification: Why Everything Suddenly Got Worse and What to Do About It. 2024. — Also see promarket.org review, December 2025.

RMIT Information Integrity Hub: 'How the Internet Drowned Itself in Slop' (December 2025). rmit.edu.au

Wired (2025). 'Trump: The First AI Slop President.' wired.com

 

PASSAGE 3  |  The Subscription Trap

Pay Forever: How Big Tech Turned Everything into a Subscription and Why You Can Never Stop

The Great Unbundling

There is a business strategy so effective, so profitable, and so quietly inescapable that it has fundamentally restructured how Americans spend money on technology — and on almost everything else. It is called the subscription model, and over the past decade it has transformed from a convenience offered by a handful of streaming services into the dominant architecture of the digital economy. Software that you used to buy once now bills you monthly. Tools you used to own are now rented. Features that were included in a product are separated out and placed behind paywalls. And because everything is connected to your data, your workflow, and your creative work, the cost of leaving is often higher than the cost of staying.

Adobe — the company behind Photoshop, the industry standard in image editing used by photographers, designers, and artists worldwide — sold Photoshop as a one-time purchase until 2013. The price was approximately $700 for a permanent license. In 2013, Adobe switched to a subscription model: Creative Cloud, which now costs $55 to $85 per month depending on the plan. A user who bought Photoshop in 1995 and used it for 30 years would have paid, in total, the cost of a few one-time upgrades. A user who subscribes to Creative Cloud for 30 years will pay between $19,800 and $30,600. If the user cancels, they lose all access to the software entirely — including files created in formats only readable by that software.

$19,800+

The minimum 30-year cost of Adobe Creative Cloud at its lowest monthly tier, compared to the one-time purchase model it replaced. Files created in proprietary Adobe formats become inaccessible if the subscription lapses.

Microsoft made the same transition. Microsoft Office — Word, Excel, PowerPoint — was for decades sold as a one-time license. Microsoft 365 now bills users $100 per year for personal use or $130 for families. The features themselves have not dramatically changed. The business model has. Microsoft's annual recurring revenue from Microsoft 365 is in the tens of billions of dollars. What changed is not the product. What changed is how long Microsoft can extract payment from the same customer.

The Lock-In Architecture

The subscription model's most powerful feature is not its monthly fee. It is what economists call "switching costs" — the real and perceived cost of leaving the platform for an alternative. Tech companies design these switching costs deliberately. They store your data in proprietary formats that are difficult to export. They build workflows, integrations, and habits around their specific tools. They make their products the default that other products connect to, so that leaving one means disrupting many. And they use AI to make their tools progressively more personalized — learning your habits, your preferences, your professional needs — so that after two years, no competing product can replicate your specific configuration.

Spotify, Netflix, Amazon Prime, iCloud, YouTube Premium, Adobe Creative Cloud, Microsoft 365, Google One, Apple One, Hulu, Disney+, Peacock, Paramount+, Apple Music, Audible, LinkedIn Premium, Duolingo Plus, and Headspace. Add them up. The average American household now spends between $900 and $1,200 per year on digital subscriptions — a figure that has more than doubled since 2018, and which many households do not accurately track because individual charges are small and easy to overlook.

A 2022 study by C+R Research found that Americans spend an average of $219 per month on subscriptions — nearly $2,600 per year — and underestimate their actual subscription spending by approximately 80%. The design of subscription billing is engineered to minimize the psychological salience of cost: small monthly charges feel much less significant than equivalent annual or one-time payments, even when the total spent is identical. Companies know this. Some companies have been found to deliberately make cancellation difficult — burying cancellation options in menus, requiring phone calls during limited hours, or making users navigate multiple confirmation screens — in order to reduce churn. The Federal Trade Commission in 2024 adopted new rules requiring that cancellation be as easy as sign-up, after widespread complaints about these practices.

AI as the Ultimate Subscription Product

Artificial intelligence tools represent the most significant new chapter in subscription capitalism. OpenAI's ChatGPT charges $20 per month for its premium tier. Anthropic's Claude Pro costs $20 per month. Google's Gemini Advanced costs $20 per month. Microsoft's Copilot is bundled into Microsoft 365 at higher tiers. These tools are not standalone products — they are portals to AI systems that learn from their users' data, making each individual's version of the tool increasingly customized over time, which increases switching costs. The more you use an AI tool, the harder it is to switch to an equivalent alternative, because the alternative starts from zero with no knowledge of your preferences, writing style, or workflows.

The scale of capital being invested in AI — estimated at $650 billion committed by Microsoft, Alphabet, Amazon, and Meta through 2026 — requires enormous revenue to justify. The subscription model is how that revenue gets extracted. An analyst report published in October 2025 estimated that big tech needs approximately $2 trillion in AI-related revenue by 2030 to recoup the infrastructure investment being made between 2024 and 2026. The pressure to reach that revenue target will be passed directly to users through subscription pricing, feature paywalling, and the progressive introduction of advertising into tools that currently charge for an ad-free experience.

The trajectory is already visible. When Netflix added an ad-supported tier in 2022 — after years of insisting it would never run ads — it was widely described as a broken promise. It was also an inevitable consequence of a business model that must continuously justify its valuation. When Spotify began running ads on its paid tier in 2023 and 2024, users reacted with outrage. The company had not promised a strictly ad-free experience at all tiers, but the precedent was clear: subscription does not mean ad-free. It means both.

Tech analyst Cory Doctorow has noted that the subscription model creates a permanent extraction relationship — what he calls "the right to keep reaching into your pocket in perpetuity." The product is never finished. The cost never ends. And because your data, your files, and your workflows live inside the platform, leaving has costs that go beyond the monthly fee.

Who Owns Your Tools?

The deeper question underneath the subscription economy is one of ownership and dependency. When you buy a hammer, you own it. When you subscribe to a tool, you rent it. The distinction matters because it determines who has power in the relationship. A renter can be evicted. A tool can be taken away, changed, or made worse. A subscription can be price-increased — and because your data lives inside it, you cannot leave without losing your work.

In 2023, Adobe attempted to add contract terms to its creative software subscription that would give Adobe the right to access, analyze, and use any content created by subscribers to train its AI systems. The backlash was significant and the terms were partially revised. But the episode revealed a key tension in the subscription model: the company providing the tool and the person using it have fundamentally different interests, and the company — which controls the terms of service, the pricing, the feature set, and the data — holds most of the power.

 

🔍  Socratic Seminar Questions

1.  Adobe's shift from a $700 one-time purchase to a subscription model that costs up to $30,600 over 30 years represents the same software. What justifies the difference in total cost? What does the scale of that difference tell you about how companies think about long-term customer relationships?

2.  Americans spend an average of $219 per month on subscriptions but underestimate that spending by 80%. Subscription companies deliberately make charges psychologically small and cancellation deliberately difficult. Is this manipulation? Is it legal? Should it be?

3.  The FTC adopted new rules in 2024 requiring that cancellation be as easy as sign-up, because companies were using confusing cancellation processes to retain customers. Why do you think it took government action to require this, rather than market competition? What does this tell you about whether the tech market is actually competitive?

4.  AI tools like ChatGPT, Claude, and Gemini charge subscription fees and train on the data users generate through their use. When you use an AI tool, who owns what you create? Who owns the interaction? Who benefits from the training data your use generates?

5.  Big tech companies need an estimated $2 trillion in AI revenue by 2030 to justify their current capital investment. That revenue will come from users. What mechanisms do you think they will use to extract it? What choices will users have about whether to participate?

6.  If every digital tool you depend on for school, work, or creative expression were taken away tomorrow because you couldn't pay a subscription fee, what would you lose? What does that vulnerability tell you about the relationship between digital access and power?

💭  Food for Thought

•  The feudal system in medieval Europe was based on the idea that land — the fundamental productive resource of the era — was not owned by the people who worked it, but rented from lords who could raise rents, change terms, and evict tenants at will. Some economists and technologists have begun using the term 'digital feudalism' to describe a world where the fundamental productive tools of the information economy are not owned by the workers who use them, but subscribed to from corporations that can raise prices, change terms, and cut off access at will. Is this comparison fair? What would need to be true for it to be accurate?

•  Adobe claimed the ability to access subscribers' files to train its AI. Your phone company stores your location data. Your email provider scans your messages for advertising targeting. Your streaming service tracks every second you watch and every second you pause. You have consented to all of this in terms of service that no one reads. If this data were being collected by a government agency rather than a private corporation, it would be described as mass surveillance. What is the meaningful difference between these two scenarios — if any?

•  The subscription model depends on inertia and switching costs. Its power comes partly from the fact that leaving is hard. What would need to be true for competition to actually discipline subscription pricing? What barriers exist to the kind of competition that would force Adobe, Microsoft, or Spotify to genuinely compete on price and user-friendliness rather than on switching costs?

•  One alternative to subscription software is open-source software — programs that are free, whose code is publicly available, and that anyone can modify or improve. GIMP is a free, open-source alternative to Photoshop. LibreOffice is a free alternative to Microsoft Office. Why do you think most users continue to pay for subscription software rather than switching to free alternatives? What does your answer tell you about the nature of the lock-in?

📚  Verify the Facts — Sources to Check

Adobe Creative Cloud pricing history and comparison to perpetual licenses — documented across multiple reviews including PCMag and The Verge.

C+R Research (2022): 'Subscription Service Study' — Americans spend $219/month on average on subscriptions and underestimate spending by ~80%. crresearch.com

Federal Trade Commission (2024): 'Click to Cancel' rule — requiring cancellation to be as easy as sign-up. ftc.gov

Campaign US (February 2026): 'Big Tech's AI Spend in 2026: Following the Money' — $650B committed; $2T revenue target. campaignlive.com

Doctorow, Cory. Enshittification: Why Everything Suddenly Got Worse and What to Do About It (2024). — Also see promarket.org, December 2025.

The Verge and Reuters (2023): Coverage of Adobe's AI training data terms controversy. theverge.com

 

PASSAGE 4  |  The Billionaire Code

Five Men, Five Platforms, Billions of Minds: The Concentration of Informational Power in the Digital Age

The Most Powerful Media Owners in History

At no point in recorded history has so much of human communication, information, and political discourse been controlled by so few people. Five men — Mark Zuckerberg (Meta), Elon Musk (X, formerly Twitter), Sundar Pichai (Alphabet/Google/YouTube), Shou Zi Chew (TikTok/ByteDance), and Sam Altman (OpenAI) — control the platforms through which the majority of Americans now receive news, discuss politics, form social connections, and understand the world. They did not run for office. They were not appointed. They are accountable to shareholders, not to the public. And the decisions they make — about what content is amplified, what content is suppressed, what speech is permitted, what information is labeled as misinformation — affect the political and social reality of billions of people.

The scale of this concentration is without historical precedent. When William Randolph Hearst controlled dozens of newspapers and magazines in the early 20th century, his reach was considered dangerous and led to widespread debate about media concentration. Hearst reached perhaps 20 million readers. Meta reaches more than 3 billion monthly active users across Facebook, Instagram, and WhatsApp. YouTube reaches 2.5 billion users. TikTok has approximately 1 billion users globally. These are not media companies in the traditional sense — they are the infrastructure through which media flows, which gives their owners a qualitatively different kind of power than any previous media baron possessed.

3.07B

Monthly active users across Meta's platforms (Facebook, Instagram, WhatsApp) as of Q4 2024 — roughly 38% of every person alive on earth, controlled by a single private company and its founder.

The Elon Musk Experiment: What Happens When One Man Buys the Public Square

When Elon Musk purchased Twitter for $44 billion in October 2022 and renamed it X, the transaction was described by many observers as the most consequential private acquisition of a communications platform in history. Twitter, despite having far fewer users than Facebook or YouTube (approximately 350 million monthly active users at the time of purchase), held an outsized role in public discourse: it was the platform used by journalists, politicians, scientists, public health officials, and world leaders to communicate information and set the agenda for mainstream media coverage.

Musk's subsequent decisions as owner of X have been extensively documented: he fired approximately 80% of the company's workforce, including most of its trust and safety team, its misinformation researchers, and its content moderation staff. He reinstated accounts that had been permanently banned for violating platform policies — including accounts associated with white nationalist content. He eliminated the verified checkmark system that distinguished public figures from impersonators and replaced it with a paid subscription (Twitter Blue/X Premium), so that any account paying $8 per month could display a verification badge regardless of who they were. He changed the algorithm to boost his own posts to the top of every user's feed. He used the platform to directly advocate for political candidates in the United States and in multiple European countries, including posting more than 100 times in support of far-right parties in Germany's 2025 elections — reaching hundreds of millions of users.

Musk's Grok AI assistant, built into X, was found to generate responses that aligned with specific political narratives rather than established facts, particularly on topics where Musk holds strong political views. Grokipedia — Grok's AI encyclopedia — was found by researchers to have copied more than half its entries from Wikipedia and significantly rewritten entries on controversial topics 'to highlight a specific narrative,' without disclosing this to readers.

Zuckerberg's Pivot: When the Fact-Checkers Were Fired

In January 2025, Meta CEO Mark Zuckerberg announced that Meta would end its third-party fact-checking program on Facebook and Instagram in the United States, replacing it with a "community notes" system modeled on X's approach. He announced the company would relax its hate speech policies and end its diversity, equity, and inclusion programs. He said the fact-checking program had resulted in "too much censorship." He announced these changes in a video posted to Facebook shortly after the presidential inauguration.

The fact-checking program, which Zuckerberg was now ending, had been implemented in 2016 after evidence that false information spread on Facebook had influenced the U.S. presidential election. Its effectiveness had been debated, but multiple studies found that labels on false content significantly reduced its spread. The timing of the announcement — immediately following a presidential transition that both Trump and Musk had been actively involved in — led many observers to describe it as a political accommodation rather than a policy evolution. Arthur Caplan, a leading bioethicist at NYU, wrote that the change was "a capitulation to political pressure from the incoming administration, not a principled free speech decision."

The consequences of removing fact-checking infrastructure from platforms used by billions of people are not abstract. During the COVID-19 pandemic, social media misinformation about vaccines contributed to lower vaccination rates and higher mortality. During election cycles, false information about voting procedures has caused people to miss voting deadlines or show up at incorrect locations. During natural disasters, misinformation about government response has caused people to refuse evacuation orders. The fact-checking program was imperfect. Its absence will not be.

The AI Race and the Democratic Deficit

The four major companies racing to dominate artificial intelligence — OpenAI (backed by Microsoft), Google/DeepMind, Meta, and xAI (Musk's AI company) — have invested a combined total that will exceed $650 billion by 2026. This investment is in the infrastructure of AI: the data centers, the energy, the computing chips, and the talent required to build and run large language models. The AI systems produced by this investment will, in their most ambitious forms, be used to write news articles, moderate political speech, generate educational content, advise on medical and legal decisions, and serve as the primary interface through which people access information.

There is no democratic process governing who builds these systems, what values they encode, what information they present as true, or what perspectives they exclude. The decisions are being made by Sam Altman, Sundar Pichai, Mark Zuckerberg, and Elon Musk — four individuals whose combined net worth exceeds $400 billion, who were not elected, who are accountable primarily to their shareholders, and whose personal political views and business interests will inevitably shape the AI systems they build.

A survey of software developers in Silicon Valley, published in a peer-reviewed journal in January 2026, found that most developers recognize their products have the power to influence civil liberties and political discourse — and that they frequently face ethical dilemmas and 'top-down pressures that can lead to design choices undermining democratic ideals.' They understand what they are building. They are building it anyway, because that is what they are paid to do.

The concentration of AI development in a small number of privately owned companies, controlled by a handful of individuals, represents what some researchers are calling a "new digital divide" — not a divide between those who have internet access and those who don't, but between those who control the information systems of the future and those who simply consume whatever those systems produce. The people consuming are billions. The people controlling are a handful. No democratic mechanism currently exists to give the billions any meaningful voice in how those systems are built, what they optimize for, or whose interests they ultimately serve.

 

🔍  Socratic Seminar Questions

1.  William Randolph Hearst's newspaper empire, which reached approximately 20 million readers, was considered dangerous enough to inspire major debates about media concentration. Meta reaches 3 billion users. Why do we have no equivalent contemporary debate — and what would that debate need to look like to be productive?

2.  Elon Musk fired 80% of Twitter's workforce — including most of its trust and safety team — and then used the platform to advocate for political candidates in multiple countries. He was not elected. He faces no regulatory accountability for the political consequences of his platform decisions. Is this a problem for democracy? What, if anything, should be done about it?

3.  Zuckerberg ended Meta's fact-checking program immediately after a presidential transition, in changes that critics described as political accommodation. He describes it as a free speech decision. How would you evaluate these two explanations? What evidence would help you decide which is more accurate?

4.  A survey of Silicon Valley developers found that most recognize their work can undermine democratic ideals, and that they face top-down pressure to build things that do. If you were a software engineer at one of these companies, what would you do when faced with a design decision you believed was harmful? What are the realistic options? What would you lose by choosing each one?

5.  There is no democratic process governing who builds AI systems, what values they encode, or what they define as true. Should there be? What would democratic AI governance look like? What would the tech industry's objections be — and how would you answer them?

6.  After reading all four passages in this series, describe the system you see connecting them. Who holds power? Who profits? Who pays the cost? And — the hardest question — what would it actually take to change it?

💭  Food for Thought

•  The First Amendment to the U.S. Constitution protects free speech from government censorship. It does not apply to private companies. This means that Facebook, YouTube, and X can legally suppress any speech they choose — or allow any speech they choose — without constitutional constraint. Some argue this means we need new laws specifically covering platform speech. Others argue that government regulation of platform speech is more dangerous than private control. Where do you come down — and what principles guide your position?

•  Historians and political scientists who study propaganda and authoritarianism have noted that one of the most effective features of authoritarian information control is not the suppression of all information, but the flooding of the information space with so much noise — misinformation, spectacle, outrage, contradiction — that citizens cannot form coherent political judgments. How does the current digital information environment compare to this model? Is it authoritarian propaganda? Is it something different? Does the distinction matter?

•  The people building the most powerful AI systems in history — and making the decisions about what values those systems embody — are overwhelmingly young, overwhelmingly male, overwhelmingly wealthy, and overwhelmingly from a narrow band of elite universities in a single geographic region. They are building tools for billions of people who had no voice in their design and no representation in the decision-making process. Is this a democracy problem? A market problem? A culture problem? Or is it an unavoidable feature of where breakthrough technology comes from?

•  This series began with the algorithm of outrage, moved through the slop economy, examined the subscription trap, and ended here — with the question of who controls the most powerful information systems ever built, and to whom they are accountable. If you were writing the conclusion of this series yourself, what would be the single most important thing you would want students to understand — and what would you want them to do with that understanding?

📚  Verify the Facts — Sources to Check

Groundy (February 2026): 'Facebook Is Cooked: Inside Meta's Platform Decay' — Meta revenue $164.5B in 2024; AI slop dynamics; platform decay thesis. groundy.com

Tandfonline (January 2026): 'A New Digital Divide? Coder Worldviews, the Slop Economy, and Democracy in the Age of AI.' Peer-reviewed. tandfonline.com/doi/full/10.1080/1369118X.2025.2566814

Wikipedia: 'Elon Musk and Twitter/X' — documented changes to platform policies, workforce reductions, and political interventions. en.wikipedia.org

Meta (January 2025): Zuckerberg announcement on ending fact-checking program. Available via Meta Newsroom. about.fb.com

Campaign US (February 2026): Big Tech AI spend — $650B committed; revenue projections. campaignlive.com

ProMarket (December 2025): Review of Doctorow's Enshittification and Wu's The Age of Extraction. promarket.org

U.S. Senate Judiciary Committee (January 2024): Hearing with CEOs of Meta, TikTok, X, Discord, Snapchat. Available via senate.gov

 

A Note on Using This Series

The Chaos Code is designed to be read alongside Greed Nation and Poisoned for Profit, as part of a larger critical literacy curriculum. Together, the three series trace a single pattern: the extraction of value from human need, human attention, and human vulnerability — and the systematic weakening of the oversight systems that might constrain it.

Students who work through all three series in sequence should be asked a final syntopical question: Across twelve passages and three topic areas, what is the one most important structural change — in law, in technology, in education, or in culture — that would most significantly disrupt the patterns described? There is no right answer. The value is in the reasoning required to develop one.

Every claim in this series is sourced. Every source is listed. Check them. The ability to verify information independently — to go beyond the feed, beyond the headline, beyond the algorithm — is not just a classroom skill. It is one of the most important capacities a person can develop in the world these passages describe.