The Silver Tsunami: Issue No. 001

Welcome to The Silver Tsunami. Every Thursday: the biggest stories in America and the lore behind them. Kevin and I break it all down on the podcast every week. Let’s get into it.

Everyone is talking about the Apple handoff: Tim Cook out after 15 years, John Ternus in, and the iPhone Duo, the $1,999 foldable Ternus unveiled this week in his first keynote as CEO. I posted the raw numbers and 2.3 million people stopped to look: Cook inherited $350 billion. He handed back $4.7 trillion. That is $4.3 trillion of market value in 15 years. About $800 million a day, weekends included. I went deep on how that actually happened, because the popular version of the story is wrong.

The lore starts in 1998, when Steve Jobs hired a supply chain executive out of Compaq named Tim Cook. Friends told Cook not to take the job. Apple was months from bankruptcy and Michael Dell had just said the company should be shut down and the money given back to shareholders. Cook took it anyway, then did something almost nobody talks about: he closed Apple’s factories. All of them. He outsourced manufacturing and cut inventory from months of product sitting on shelves down to days. His line about it became famous: inventory is fundamentally evil. You manage it like you are in the dairy business, because it spoils. That boring dairy logic is what funded every hit you remember. The iPhone was a miracle of design, but Apple captured the profit because Cook’s supply chain could build the miracle at scale, in secret, at margins nobody could touch.

Then August 2011. A dying Jobs hands him the company and tells him one thing: never ask what Steve would do. Just do what is right. Wall Street shrugged. The caretaker thesis was everywhere: the visionary is gone, the ops guy will manage the decline. The decline turned out to be a 13x.

And here is the part getting the least coverage: Apple has been running its succession machine the entire time. Executives rotated through divisions to broaden them. John Ternus, the hardware engineering chief, positioned publicly for years in keynotes, then announced back in April with a five-month runway before taking the chair on September 1. By the time the handoff happened, the board was not choosing a successor. It was ratifying one. And Ternus’s first week as CEO was not a crisis press conference. It was walking on stage with the iPhone Duo. The transition was engineered to look like a product launch, because it was one.

Compare that to Disney, which botched succession so badly it had to drag Bob Iger back out of retirement. Succession is the hardest problem in corporate America, and the track record even among giants is ugly.

Now the angle I cannot unsee. Apple spent more than a decade preparing one job handoff. Meanwhile the average American business owner is in their 60s, and most have no written succession plan at all. Not a decade. Not a page.

Both sides, because they exist: skeptics say Cook’s real test was never passed, that he optimized a machine Jobs built, and that the next guy inherits a company that has not shipped a category-defining product in years. Fair. Ask again in five.

The best succession stories are boring on purpose. Boring is what a decade of planning buys you. Everything else is a Disney sequel.

We asked Claude, ChatGPT and Grok the same question: does Apple’s next CEO need to be a product visionary, or is the visionary era of tech CEOs over?

Claude: Yes, Apple's next CEO needs to be a product visionary. Apple's premium pricing rests entirely on product taste, so a pure operator turns it into a commodity hardware business with a great installed base. The next decade is an interface shift, and those get won by someone willing to cannibalize the current business before a rival does, which is a judgment call committees never make. The visionary era isn't over generally, it just requires the board to grant a hired CEO the authority to be wrong expensively, which is the open question at Apple.

ChatGPT: No. Apple’s next CEO after John Ternus does not need to be a Steve Jobs-style product visionary; the visionary era isn’t over, but the singular visionary CEO era largely is. Modern tech is too complex for one person to invent the future—the CEO’s job is increasingly to recognize great product bets, concentrate resources behind them, and build an organization where exceptional product people win. Apple still needs taste and product conviction at the top, but it needs an editor-in-chief more than an auteur.

Grok: Apple still needs a product-oriented CEO. The lone-genius visionary era is over. Ternus was chosen because Apple’s value still comes from integrated hardware that people actually want to hold, not from operations or services alone. Cook already proved a brilliant operator can scale the company; the current job is deciding which new form factors and on-device AI experiences are worth shipping. At this size, the winning model is a product-literate operator, not another Steve Jobs.

We’ll see who’s right!

The fastest $500 I have ever spent, and 2.1 million of you watched me spend it. But the toothbrush is not the story. The company that could build it is.

Dyson spent six years, 661 engineers, and 38 patents on a toothbrush: a 100,000-pixel camera scanning your teeth at 28 images a second, firing a 100 millisecond jet of water at what it finds, removing 69 percent more plaque. Everyone asked the same question: who spends that on a toothbrush?

Here is who. A private, family-owned company. That is the lore worth knowing. James Dyson built 5,127 prototypes of his first vacuum over five years while going broke. Every major manufacturer rejected it, because the bagless vacuum threatened their replacement bag revenue, a detail that should sound familiar to anyone who has watched an incumbent protect a margin. So he launched in Japan, licensed the design, and used the royalties to build his own company. He never sold it and never took it public. Today Dyson is a multibillion pound family business, and James Dyson personally became one of the largest private farmland owners in England.

A public company cannot spend six years on a toothbrush. The quarterly call will not allow it. A founder-owned company answers to no one, which is why the most over-engineered products in your house keep coming from the same few names. Ownership structure is product strategy. Readers of this newsletter already know the punchline: the people who own the asset outright are playing a different game than the people managing someone else’s.

$499. But you cannot put a price on health.

The entire legal profession is having a “what comes next” moment, and this week the clients started saying the quiet part out loud: if AI does the work in minutes, why does the bill still say hours?

The thesis: the billable hour survived every technology of the last seventy years because clients could not see inside the work. AI is the first one that lets them see.

The lore, because almost nobody knows where the billable hour comes from: lawyers did not always sell time. For most of American history they charged flat fees and contingencies. The hour took over in the 1950s, pushed by bar association studies with titles like “The 1958 Lawyer and His 1938 Dollar,” which told attorneys they were undercharging and that tracking time was the cure. It worked so well that within a generation the industry forgot any other model existed. Rates at elite firms now run north of $2,500 an hour for top partners.

The problem was always the incentive: hourly billing pays for effort, not outcomes. Inefficiency is revenue. Every general counsel has known this forever, but they could never price the alternative because they could not see how long the work should take. Now they can. When a first-year associate’s twelve hours of diligence review becomes a twenty-minute AI pass plus review, the bank’s procurement team does not renegotiate. It re-prices the entire relationship.

Full disclosure: I run a law firm and we bill hourly. I am not writing this from a rooftop, I am writing it from inside the building. Here is what I think the banks have right and the doom headlines have wrong: the hour is not dying because lawyers are greedy. It is dying because it was always a proxy for what clients actually wanted to buy, which is judgment plus outcome. AI strips out the proxy. The firms that know their true cost of production will thrive on fixed pricing. The firms that only ever knew how to sell time are about to find out they never knew what anything cost.

Who gets rich: firms with real process, real data on their own economics, and the nerve to price outcomes. Who gets crushed: the middle of the market, where the leverage pyramid of junior hours was the whole business model.

The billable hour had a seventy-year run. The clients just got the one thing they never had: a comparison.

An AI company warning that AI might end humanity is either the most honest thing in tech or the greatest marketing in history, and this week Anthropic had everyone arguing about which

The lore: Anthropic was founded in 2021 by a brother and sister, Dario and Daniela Amodei, who walked out of OpenAI with a group of researchers over how fast to push the technology. The premise was safety first, the structure a public benefit corporation, and its CEO has publicly put the odds of catastrophic outcomes at double digit percentages while raising billions to build the very thing he is worried about. Critics call that a contradiction. The company’s answer is the seatbelt argument: the cars are getting built either way, so the people worried about crashes should be the ones building them.

Both sides, laid out straight. The cynical read: doom talk is a moat. Scare regulators into licensing regimes and only the giants can afford compliance, the warnings are self-serving, and nobody who truly believed the end was near would ship a subscription product. The sincere read: the engineers closest to a technology have warned first before, the company publishes safety research that sometimes makes its own products look worse, and “they must not mean it” is not an argument, it is a coping mechanism.

My operator take: watch what the labs do, not what they say. Watch what they pay for, what they refuse to ship, and what they lobby for in the fine print. Every industry I work in prices risk in documents nobody reads. This one is pricing it in press releases everybody reads and nobody believes.

This week’s Split, and it is the best one we will ever run: we asked Claude, ChatGPT and Grok whether an AI company warning about AI should be believed. Yes, we asked Claude about its own maker.

Claude: Judge the claim, not the messenger. An AI company's warning carries real information because the people building the systems see capabilities first, but it also runs on mixed incentives: doom talk sells capability, and heavy regulation tends to favor incumbents who can absorb the compliance cost. So treat it as evidence worth taking seriously, discount it where the warning conveniently benefits the company, and weight it against independent researchers who have no equity in the outcome.

ChatGPT: Yes—but not uncritically. An AI company has unusually strong access to technical evidence about its own systems, so its warnings deserve serious weight, even though it may also have commercial, regulatory, or competitive incentives that shape how it frames the risk. The rational approach is to treat the warning as important evidence, then ask whether independent researchers and observable facts support it.

Grok: No. AI companies have strong incentives to overstate risks (regulation that favors incumbents, funding, or publicity), so their warnings should be treated as interested claims rather than disinterested expertise. Judge the specific arguments and evidence on their own merits.

I don’t believe any of them!

Two college football stories went viral on my feed this week, and they are the same story wearing different helmets: the sport’s money has outgrown its governance.

Story one you already know if you follow me: Michigan beat Western Michigan on a play that only existed because a replay booth used the wrong clock. The MAC protested, an NCAA memo agreed the protocol was violated, and nothing changed, because no tribunal in college football has the power to reverse a final score. The MAC’s appeal literally asked the playoff committee to “recognize” the true winner. The remedy for a proven error in a multibillion dollar sport is a memo. Disclosure: Western Michigan is my school in Kalamazoo. Bias declared.

Story two: LSU. I am the last person to defend LSU, but the sight of that much legal machinery deployed over two guys who never played an NFL down tells you what the sport has become. When the TV money hit billions a year and NIL turned rosters into payrolls, every dispute stopped being a sports story and became a commercial dispute with a fight song.

The thesis: college football is a multibillion dollar industry still governed like an intramural league, and the gap between the money and the governance is where every one of these stories comes from. The Big Ten’s media deal alone runs roughly $7 billion. A playoff berth is worth tens of millions to a conference. Which means one second on a clock in Ann Arbor is not a second. It is a line item. The lawyers did not invade college football. The money sent for us.

One more from my feed, handled with the sobriety it deserves. The Lindsay Clancy trial went to the jury, and my posts tracking the deliberation deadlocked 11 to 1 after seven days and 36-plus hours; verify current status and update before publish drew enormous attention, because the case sits on the hardest question in criminal law: what does the law do when the person who committed the act may not have been the person, in any meaningful sense, at the time.

The legal lore, for everyone arguing about it online: the insanity defense is not a modern invention and not a loophole. The modern test dates to 1843, when a Scottish woodturner named Daniel M’Naghten, in a delusional state, killed the British Prime Minister’s secretary. The outrage at his acquittal produced the M’Naghten rules: did the defendant know what he was doing, and know it was wrong. Massachusetts uses a broader standard: whether mental disease left the defendant without substantial capacity to appreciate the wrongfulness of the conduct or to conform it to the law. Two facts that surprise everyone: the defense is raised in less than 1 percent of felony cases and fails most of the time it is raised, and a verdict of not guilty by reason of insanity typically means indefinite psychiatric commitment, often longer than a prison sentence. The closest parallel is Andrea Yates: convicted in 2002, verdict overturned, found not guilty by reason of insanity on retrial in 2006 as the science on postpartum psychosis got its day in court.

I said it on X and will say it here: explaining why this case is legally hard is not legitimizing every take about it. A deadlocked jury is not a broken system. It is twelve people taking the hardest question in the law seriously.

While Apple spent a decade on one succession: roughly 2.9 million US businesses are owned by boomers, and only a small fraction that go to market ever sell. The rest just close. That transfer is the biggest money story in America, and it is why this publication has its name.

John Ternus. Joined Apple’s product design group in 2001 and spent two decades inside the hardware machine: iPads, iPhones, the Mac’s chip transition, eventually running all of hardware engineering. Then came the tell: year after year of growing keynote time, the classic sign of a board grooming its pick in public. The experts wanted a visionary. Apple picked the engineer who ships, gave him a five-month runway after the April announcement, and handed him the iPhone Duo to walk out with in week one. The company does not chase heroes. It builds systems that make the next person’s job survivable. That is kind of the whole newsletter in one hire.

See you next Thursday.

Eric and Kevin