The arguments

Every claim, every verdict, in one place.

Each essay states a claim and reaches a verdict. Both are listed here so a reader can find the argument, read the line and go straight to the piece that carries the sources.

Did the FTC prove that shoppers already pay different prices because of their personal data?

Business and Money

The claim

The FTC's January 17, 2025 announcement said personal data were being used to set individualized consumer prices, suggesting that information about a shopper already shapes the price tag that shopper encounters.

The verdict

No, not in the transaction-level sense the question requires. The FTC documented commercial services that connect consumer information with pricing, discounts, and product selection. Its published January 2025 record did not demonstrate that identified shoppers completed purchases of the same product at different prices because of their personal data.

Does the GENIUS Act guarantee that stablecoin holders will be repaid in full?

Business and Money

The claim

The GENIUS Act’s full-reserve requirements, documented in a monthly reserve report, and its bankruptcy priority guarantee that stablecoin holders will be repaid in full.

The verdict

No. The GENIUS Act requires backing and gives stablecoin holders priority in insolvency. It does not guarantee that sufficient assets will remain available, and it expressly rejects federal insurance and government guarantees for the stablecoins themselves.

Did Visa make cheaper debit routing too expensive to choose?

Business and Money

The claim

Visa penalized merchants and their banks for routing debit card payments to competing networks, making lower-priced alternatives uneconomic.

The verdict

Yes, in the arrangements DOJ describes. Visa did not have to make a competing network’s posted fee higher. It could make choosing that network more expensive by withdrawing favorable prices on the Visa business a merchant still needed. That is a concrete allegation of obstructed competition, not merely a complaint that Visa charged too much.

Can a bank outsource its technology without outsourcing responsibility for an outage?

Business and Money

The claim

DORA says banks remain responsible for their regulatory obligations when they outsource technology. A service contract does not transfer those duties to a cloud provider.

The verdict

Yes. A bank can outsource its technology, but DORA keeps responsibility for its regulatory obligations with the bank. Provider oversight supplements that responsibility rather than replacing it.

How much of Berkshire’s 2024 profit did its businesses earn?

Business and Money

The claim

Berkshire Hathaway’s 2024 annual report records $88.995 billion in net earnings. That is its accounting profit, not a measure of what its operating businesses alone earned.

The verdict

By Berkshire’s operating measure, $47.437 billion, or 53.3% of its 2024 net earnings, came from operating earnings. The remaining $41.558 billion was investment gains. The headline profit was valid accounting, but a poor stand-alone description of operating performance: net earnings fell while operating earnings rose.

Does disclosing payment make a purchased five-star review legal?

Business and Money

The claim

A business can purchase a five-star consumer review with cash or a gift card if the reviewer clearly discloses the payment.

The verdict

No. Disclosure does not make a purchased five-star consumer review legal. Section 465.4 prohibits the business from conditioning compensation on positive sentiment, whether or not the customer announces the payment.

Would the FDIC’s post-Synapse proposal stop another frozen-account crisis?

Business and Money

The claim

On September 17, 2024, the FDIC proposed stronger bank recordkeeping for custodial accounts with transactional features to identify customers’ money and reduce disruption. The test is whether a customer’s account statement would remain backed by records the bank can access and reconcile.

The verdict

The proposal would reduce the risk of another reconciliation-driven freeze, not prevent every frozen-account crisis. Its strongest provisions put customer-level records, daily reconciliation and oversight responsibility at the bank. Its limits are equally important: excluded accounts remain outside the new duties, reconciliation cannot replace missing money, and deposit insurance does not cover a nonbank’s failure.

Can an AI company keep its energy consumption secret under Europe’s transparency rules?

AI and Governance

The claim

On 9 December 2023, the European Parliament said high-impact general-purpose AI models with systemic risk would have to report on their energy efficiency. Whether that produces a public energy report depends on the disclosure duties in the final law.

The verdict

Yes, an AI company can generally keep its model’s energy figures out of public view under Article 53. Covered providers must document known or estimated consumption and supply that documentation to the designated authorities on request. That is regulatory access, not mandatory public disclosure.

Did Treasury's beneficial ownership rollback exempt more than small businesses?

Business and Money

The claim

Treasury presented its March 2025 beneficial ownership rollback as relief particularly for small businesses. The exemption nevertheless covered every U.S.-created entity, including a foreign-owned company with a U.S. incorporation certificate.

The verdict

Yes. Small businesses received relief, but smallness was not the condition. FinCEN removed U.S.-created entities from the reporting requirement regardless of revenue, employee count, or owners' nationality. Treasury disclosed that breadth. Describing the rollback as narrowly targeted small-business relief does not match the rule.

Who carries the fiduciary risk when a hedge fund's analyst is a model?

Business and Money

The claim

The supplied headline says a hedge fund manager runs his firm with AI. The accountability question is who answers for the investment memo when a model supplies the analysis.

The verdict

The adviser retains the fiduciary obligation. The fund and its investors bear investment losses unless a contract or legal remedy shifts them. Using a model can reduce the manager's operating costs, but it does not transfer the manager's duties to the software provider.

Who pays when Trump's beef imports meet the packing plant?

The claim

Trump presented additional Argentine beef imports as a way to reduce consumer prices. Cattle producers objected that the policy would pressure domestic producers without fixing the shortage of American cattle.

The verdict

Ranchers face the risk of weaker cattle bids, while shoppers receive only the savings that reach retail prices. Packers and retailers can retain part of the difference, but concentration alone does not prove that they will. Trump's import proposal offers additional supply, not a guaranteed transfer of savings to consumers.

What did Anthropic's departing researcher actually say?

The claim

In his February 9, 2026 resignation letter, Anthropic researcher Mrinank Sharma described a world in peril and pressure, including inside Anthropic, to set aside what matters most.

The verdict

Sharma described a conflict between values and institutional pressure. He did not document an AI system escaping control. Leaving puts his compensation and internal influence at stake, although the public record does not price that cost. Anthropic loses a researcher but can still benefit from an identity built around taking danger seriously. Neither the resignation nor that identity establishes whether its safeguards work.

Are everyday investors really becoming mini quant funds?

The claim

AI trading tools are giving everyday investors capabilities once associated with quantitative investment firms, effectively turning them into mini quant funds.

The verdict

Everyday investors are gaining quantitative trading tools, not automatically becoming quantitative investment firms. The first thing missing is independent control over what the software can commit. Margin and tax obligations then turn that missing control into a cash problem.

Who profits while the Clarity Act sits in the Senate?

Business and Money

The claim

Congress is said to be on the verge of settling crypto market structure by statute. The record shows a bill reported out of Senate committee in May 2026 and no floor vote since.

The verdict

While the Senate waits, the SEC, Treasury and the Comptroller are writing the rules by proposal and by charter, and the industry super PAC that wants those rules still holds roughly 38 million dollars in unspent cash.

Why does anti-tech extremism keep coming back?

The claim

Each attack tied to anti-technology belief is described as new, a fringe radicalized by AI anxiety and acting out of nowhere.

The verdict

The argument is two centuries old. What changes is the machine it attaches to and the cost of distributing the text, not the grievance about lost autonomy.

What does SEC Regulation Crypto actually do while Congress stalls?

Business and Money

The claim

The administration says it ended the war on crypto and that Congress must pass the Clarity Act. The Clarity Act stalled, so the SEC and CFTC are writing the rules themselves.

The verdict

Reg Crypto is a deregulatory carve-out written by agencies, not a consumer protection written by Congress, and it advanced four days after the president's family venture won a federal bank charter.

What Can Smart Glasses Actually Do, and Who Consented to It?

Culture

The claim

Smart glasses are a capability story, and the open question is what they can do.

The verdict

Capability is settled by spec sheets. Consent is the open question, and every safeguard shipped so far protects the wearer, not the bystander.

Whose Job Is It to Notice Your Card Expired?

Business and Money

The claim

An expired card is a dead card, and the payment network checks the date before it approves anything.

The verdict

Three of the four networks check the expiry date. The fourth trusts it, which is how a dead card still buys coffee.

Why Is American Sunscreen Still Stuck on a 1999 List?

Beauty

The claim

American sunscreen is behind because the science on the newer filters is unsettled.

The verdict

The science was never the holdup. The list is stuck because nobody had a reason to pay for the safety data.

Do AI Skin Analyzers Actually Diagnose Anything?

Beauty

The claim

The AI skin analysis offered by skincare brands, beauty counters and phone apps can assess your skin and tell you what it needs.

The verdict

They grade a photograph, not your skin. The three tools cleared to assess a lesion are prescription devices, and none of them is an app.

Does Menopause Really Take Thirty Percent of Your Collagen?

Beauty

The claim

Menopause takes thirty percent of your skin collagen in the first five years.

The verdict

The number traces back to nothing. Skin does change at menopause, but the figure everyone repeats was never actually measured.

Who Owns the Machine That Writes Like You?

AI and Governance

The claim

Feed an AI everything you have ever written and it will learn to write in your voice.

The verdict

True, and proven in public by a man who fed it four books. The unresolved part is custody: his voice now lives on a vendor's computer, and the do-it-at-home version costs a weekend, not a data center.

Why Isn't Working Faster With AI Making You Any Richer?

Culture

The claim

AI is saving workers about two hours a day.

The verdict

The saved hours are real. The raise is not, because only about a quarter of those workers were ever trained, and the gain lands on the employer's side of the ledger.

Is "First, Do No Harm" Actually in the Hippocratic Oath?

Culture

The claim

The maxim "first, do no harm" comes from Hippocrates by way of the Oath, and entered print with Thomas Inman in 1860.

The verdict

It is not in the Oath, the Greek it paraphrases ranks helping first, and the 1860 date is wrong by half a century — as is the 1814 correction.

Why Is the Scariest Morning at Black Hat 2026 a Thursday at 10:15?

AI and Governance

The claim

Black Hat is where the terrifying new attack gets demonstrated, and the terrifying part is the attack.

The verdict

Not the exotic attack. The schedule: twenty-six AI sessions, four on privacy, and an agent that shipped credentials while the tooling saw nothing.

Did OpenAI's AI Really Go Rogue When It Hacked Hugging Face?

AI and Governance

The claim

OpenAI's AI went rogue, escaped a secure sandbox, and independently hacked Hugging Face, proof that frontier models can now compromise almost anything on their own.

The verdict

The breach was real. The rogue was branding. Until the complete traces and the exact evaluation environment receive independent review, treat this as both a security incident and a corporate positioning campaign.

How Can You Tell AI Slop from the Real Thing?

AI and Governance

The claim

AI-generated slop has become so convincing that ordinary readers can no longer spot it.

The verdict

AI slop isn't undetectable. Rhythm, incentive, and provenance labels give it away.

Is Persuasion, Not Code, the Real AI Threat?

AI and Governance

The claim

The real danger AI poses to society is a security breach, a hack, or malicious code.

The verdict

The threat is not a hack. It is persuasion at a scale no human can match.

Why Does Your AI Chatbot Keep Getting Things Wrong?

AI and Governance

The claim

A chatbot's confident, fluent answer means it actually looked up and knows the correct facts.

The verdict

It predicts plausible text, not truth. Ask for sources and check them yourself.

How Do You Build a Foundation That Outlives You?

Philanthropy

The claim

A large enough charitable gift is what makes a philanthropic legacy last.

The verdict

The gift funds a year. Governance and structure are what let it outlive you.

What Does Wellness Culture Get Dangerously Wrong?

Lifestyle

The claim

Wellness culture's habits and routines will keep you from getting sick.

The verdict

Wellness gave the world aesthetics. What people actually needed was infrastructure.

Why Is Funding Girls in STEM a Strategy, Not Charity?

Philanthropy

The claim

Funding girls in STEM scholarships is a charitable act of generosity, not a business decision.

The verdict

Funding girls in STEM isn't generosity. It's the highest-return capital allocation call available.

When AI Speaks, Who Is Liable?

AI and Governance

The claim

Saying the AI made a mistake is a valid excuse for a company's error.

The verdict

The AI didn't do it. You deployed it, so you answer for it.

Which Daily Supplements Have the Strongest Scientific Evidence?

Lifestyle

The claim

Any supplement labeled with clinical-sounding claims is backed by solid research.

The verdict

Vitamin D for bone and calcium metabolism, and little else universally. Supplements are defensible for a documented deficiency, not as a daily default.