The Machine
That Changed
the World

On the morning of July 14, 2026, IBM opened its earnings call expecting to report a solid quarter. What followed was a rout. Shares fell more than 22% by market close — the steepest single-session decline in IBM's 115-year history. The only comparable moment was Black Monday in October 1987, when the whole market collapsed and IBM fell with it. This time, it fell alone.

CEO Arvind Krishna delivered the verdict bluntly: clients had redirected IT budgets toward AI hardware — toward Nvidia, toward TSMC-built silicon, toward the infrastructure layer of the AI boom. IBM, selling software, services, and hybrid cloud consulting, found itself in a familiar position: present at the revolution, but somehow adjacent to where the money was flowing.

The miss wasn't catastrophic in absolute terms. It was the explanation that the market found devastating. When your CEO explains that enterprise clients are moving budgets away from your core product categories and toward infrastructure you don't make — that's not a bad quarter. That's a strategic crisis wearing quarterly clothes.

IBM has been here before. It's the recurring nightmare of a company that keeps arriving at technological transitions with the right answer to the wrong question.

To understand what IBM lost, you have to understand what IBM was. In the early 1980s, buying IBM wasn't a technology decision — it was a career protection strategy. If the project failed, it couldn't be your fault. You bought IBM. You did the right thing.

The foundation was the System/360, launched in 1964. IBM bet the entire company on a single, compatible family of computers — a commitment that cost more in real terms than the Manhattan Project. Thomas Watson Jr. later called it "the biggest, riskiest decision I ever made." It worked so completely that it redefined what a computer company could be.

The trajectory is a masterclass in corporate longevity and strategic dislocation simultaneously. IBM didn't collapse like Lehman Brothers, in a sudden weekend. It declined over decades — always present, always respectable, always slightly repositioning toward the next wave while the wave moved faster than the pivot.

Paul has put his finger on exactly the right thing. In 1980, IBM needed to ship a personal computer fast. They were late to the market Apple had pioneered. Their solution was elegant in its urgency and catastrophic in its long-term implications: use off-the-shelf components instead of proprietary ones.

Intel made the chips. Microsoft licensed the operating system (from a small company called Seattle Computer Products for $75,000, then immediately licensed it to IBM). The architecture was published openly. Any manufacturer could build an "IBM Compatible" machine that ran the same software.

The result was commoditisation. By the late 1980s, you could buy a Compaq, a Dell, a Gateway, an AST for a fraction of IBM's price, running identical software with equivalent performance. The question "why pay the IBM premium?" had only one honest answer: brand reassurance. And brand reassurance alone cannot sustain a hardware business.

Andrius's reframe is genuinely important. George Stephenson and Robert Stephenson built the first steam locomotives and the first railway lines. British engineers essentially gifted railroad technology to the world. The knowledge left Britain. The benefit — global industrialisation, trade networks, food security — was civilisational. The Brits moved on to finance, insurance, and empire.

Tim Berners-Lee created the World Wide Web at CERN in 1989 and deliberately refused to patent it. The open web he created is now worth trillions of dollars — none of which went to him personally. He made that choice consciously, in full knowledge of its implications.

IBM's open architecture wasn't an act of civilisational generosity. It was competitive impatience — and the window to close it had passed before they understood what they'd opened. Apple was watching carefully. They drew the opposite conclusion.

Steve Sasson was a 24-year-old Kodak engineer when he built the first digital camera in 1975. It weighed 8 pounds, took 23 seconds to capture a single black-and-white image, and stored it on a cassette tape. Resolution: 0.01 megapixels. He presented it to management with the obvious implication: this is the future of photography.

The response, as Sasson later described it, was polite and final: "That's cute, Steve. But don't tell anyone about this." The camera was quietly shelved. Kodak did eventually invest in digital — $2 billion over the following decade — but always as a defensive move, never as a conviction bet. They were protecting the annuity, not building the future.

Clayton Christensen published The Innovator's Dilemma in 1997, a year after Kodak's revenue peak. His diagnosis fit Kodak perfectly: a company with strong financials, excellent management, and loyal customers that still failed — because it rationally optimised for its existing business at the expense of a disruptive technology that was initially worse on every metric its customers cared about, and then suddenly better on all of them.

The cruelest detail: Kodak's digital photography patents were worth billions in the smartphone era. Samsung and Apple both licensed Kodak IP. The inventor of digital photography died in Chapter 11, collecting royalties from an industry it accidentally created, forty years too late to capitalise on it properly.