Essay Series
A collection of essays on building systems that last, the infrastructure behind intelligent companies, and the thinking that shapes how work gets done. New issues published regularly.
The greatest infrastructure advantages rarely looked like infrastructure at the time, they looked like plumbing. This feature essay argues that AI is becoming the cable every company depends on, and the question is no longer whether you use it, but who owns it.
In 1928, a woman found a snail in a bottle of ginger beer. The case that followed produced a foundational principle of modern negligence law. Nearly a century later, it still shapes how courts think about responsibility. This essay asks why businesses, unlike common law courts, almost never preserve the reasoning behind their own decisions, and what changes when they finally do.
Organisational forgetting was never really about information vanishing, it was about the connection between what a company already knows and the moment that knowledge matters quietly eroding. This essay uses the tsunami stones of northern Japan to argue that the real transformation underway is not faster retrieval, but history that surfaces on its own, at the exact moment a decision is being made.
For most of business history, the honest answer to whether a company was really one organisation, or simply a great many individuals standing near each other and sharing a name, was closer to the second description than the first. This essay argues that the most consequential change now underway is not a new tool being added to the assembly, but the assembly itself beginning, for the first time, to become something more coherent than the sum of the people inside it.
A mechanical system with fifty points of contact, each losing a modest two percent of its energy to friction, delivers not ninety-eight percent of its original power but closer to thirty-six, fifty individually negligible losses compounding into one too large to ignore. Companies run on the same mathematics: thousands of small daily points of friction, searching, waiting, repeating, that no single measurement ever captures. This essay argues that the most valuable intervention in organisational productivity is not adding force or labour, but removing resistance, the organisational equivalent of Sven Wingquist's self-aligning ball bearing.
For most of history, a ruler was expected to personally understand everything the state did. The cabinet changed that. Now something structurally similar is becoming available to every employee inside a company, not merely to the executives at the top of it.
Every economic revolution solved one scarcity only to reveal another. The agricultural revolution solved the scarcity of calories, only to reveal the scarcity of allocation. The industrial revolution solved the scarcity of production, only to reveal the scarcity of distribution. Information abundance created a poverty of attention. Now, for the first time, the capacity to turn information into sound judgment, intelligence itself, is becoming the scarce resource that determines who thrives.
A city is not a machine to be optimised for throughput. This essay argues that intelligent enterprises are not built by installing isolated AI tools, but by redesigning the relationships, workflows, and decisions that make the whole organisation function.
In the 1790s, Gaspard de Prony organised thinking into a hierarchy for the first time. For two centuries, every tier required a person. AI now occupies one of those rungs, and the question facing every organisation is not how to adopt a new tool, but how to redesign the hierarchy itself.
When Benjamin Franklin left money to grow untouched for two centuries, it became millions. The same math applies to organisational knowledge. On the difference between spending a return the moment it arrives, and leaving it to compound into something considerably larger.
Nobody in a modern office building has ever paused to admire the water pressure. The technologies that changed civilization most completely are the ones we stopped talking about. This essay argues that AI is heading for the same fate, and that is the highest compliment it can receive.
A company has always had one kind of memory: the fragile, individual, endlessly leaking kind. This issue argues that the next great organisational shift is building the second brain that lets experience consolidate across the whole business.
When NASA went to rebuild the F-1 engine decades later, it had the original drawings. What it had lost was the judgment behind them. The same pattern plays out in every growing company, invisibly, expensively, and almost never noticed until the cost has already been paid.
A clock knows nothing. A body adapts and remembers. The distinction between a mechanism that repeats and an organism that learns is the one most executives have not yet drawn about their own companies.
For thirty years after electrification began, factory productivity barely moved. The gains came only when companies redesigned the factory itself. Executives adopting AI as a faster tool today are repeating the same mistake, and missing the same far larger reward.
The grand hotels of the nineteenth century solved a problem that had nothing to do with rooms. They hired a concierge so a guest never had to search. Corporate computing has spent a century asking employees to behave like a guest without one.
In 1956 the shipping container turned a fragmented industry into one interoperable system. Businesses are running their software the way global shipping ran before the container, a stack of excellent, isolated tools, none able to hand information to the next without a human repacking it by hand.
The economist Hernando de Soto showed that the world's poor were rich in assets they could not use, dead capital, unconnected to any system of record. Most companies are sitting on the exact same problem, hidden in an archive of proposals, contracts, and notes that almost nobody can find when it matters.
In 1841 two trains collided and the org chart was invented. Executives now face a comparable inflection point: what does an organisation look like once part of its workforce is not human, and what kind of leadership does that require?
Companies spent a decade building data lakes that centralised everything and clarified nothing. The Rosetta Stone sat unread for twenty-three years, the lesson is that storage was never the problem, relationship was. This essay explains why Company Brains win where data lakes didn't.
For two centuries, chemists have tried to reproduce Stradivari's varnish, and failed. The secret was never the formula; it was a lifetime of judgment that died with him. The same pattern plays out inside companies every time a long-tenured employee walks out the door.
For fifty years, software meant a product built by engineers and sold to businesses. That definition is quietly becoming obsolete. What happens when a company can turn its own accumulated judgment into something operational, without hiring a single developer?
Every company keeps a balance sheet. Nobody tracks what the organisation actually knows, or what it loses when someone walks out the door. This essay examines why institutional memory is the most undervalued asset in business, and why the companies that preserve it will quietly stop making the same mistake twice.