Chapter 5: Cyrus Smith Builds a Civilization

What this chapter does. Bootstrap leadership as a weapon against corporate bureaucracy. How the engineering team of The Mysterious Island became the prototype of every successful tiger team — from Skunk Works to the first Macintosh crews — and why the companies that build "islands" outrun the ones that build "cruise liners."

Amiens, 1873

Jules Verne has just closed the loudest success of his career. Around the World in Eighty Days has thundered through the pages of Le Temps as a serial and has sold out as a book. Hetzel demands a sequel. But Verne is tired of "planetary tourism." Fogg circled the globe on ready-made routes — the railways and steamship lines built by other people. The next book will be about the opposite. What happens if you strip away everything ready-made.

"This will be a novel about chemistry," Verne writes to Hetzel. While the public waits for new adventures, he walks the chemical factories and interrogates the engineers. How to smelt iron from bog ore. How to obtain nitroglycerin. How to fire glass. He builds the novel the same way his heroes will build a civilization — from raw materials and from the knowledge in their heads.

The idea reduces to one line. An island. Five men. No one else. No machines, no ready infrastructure, no procedures. Verne does not yet know that he is inventing the operating model of every future tiger team — from Skunk Works to the first Macintosh crews.

On the island there are no meetings.

Somewhere in the southern Pacific lies an island that does not appear on the maps. On this island, five men do what modern corporations spend armies of consultants and years of reorganization on: they build a civilization from scratch. Without procedures. Without approvals. Without alignment committees.

The team of the island — an engineer, a journalist, a sailor, an inquisitive teenager, and a servant. Five people and one dog. Nobody went through the six-round interview process that runs three to six months in modern corporations. Nobody filled out a personality inventory. Nobody sat through the "culture interview," in which an HR manager with a psychology degree decides whether your local flavor of neuroticism suits the company. Cyrus Smith simply knows who is capable of what, because the team's survival depends on competence — not on compatibility with the corporate values on the poster.

The team has no job titles. It has functions. It has no grades. It has tasks that need to be solved today so that tomorrow no one dies. It has no quarterly performance review. It has nature, which will kill you for a mistake faster than any manager can write a performance improvement plan.

Jules Verne serialized The Mysterious Island in 1874–1875, a year after the book edition of Around the World in Eighty Days. If Fogg from our opening chapter showed how one man applies an iterative approach, Smith shows something more fundamental: how to build a team that does not need a methodology, because it is a methodology. Everyone understands why. Everyone knows how. And between the "why" and the "how" — zero intermediate layers.

Not because Smith is a dictator. Because when you need to smelt iron from bog ore, turn it into tools, and use those tools to build a bridge in three days — without Google, without Stack Overflow, without a contractor, without an Agile coach who would explain to the team the importance of the daily stand-up — bureaucracy is not merely inefficient. It is lethal.

The previous four chapters were autopsy reports. Verne the publisher showed how to create the system of serial production his descendants would turn into an assembly line (Chapter 1). Shelley: how creators abandon their creations because they are afraid of responsibility (Chapter 2). Holmes: how to recognize the agony of an organization by the symptoms it carefully hides from the outside world but demonstrates with every corporate decision (Chapter 3). Borges: how measurements devour what they measure, until the map becomes more important than the territory (Chapter 4).

Act I ended with the diagnosis. Act II begins with the treatment.

And the first medicine is the most radical. Not to reform management. Not to "make it flatter." Not to "optimize the processes." Remove it. Wholesale. Replace it with what Cyrus Smith built on the island — a team that manages itself because each of its members is competent enough not to need a supervisor.

We return once more to Verne, but now to Verne the organizer, not Verne the publisher. To the man who in The Mysterious Island described a team architecture so precisely that the engineers of Lockheed would copy it seventy years later, not knowing they were copying a French novel.

The thesis of this chapter: the best projects in the history of technology have worked as autonomous islands, not as cruise liners. Small bootstrap teams with an engineer-leader outrun armies of managers with presentations.

Middle Management as an Unnecessary Organ

Before we go further — a claim that will enrage everyone who wears the title "Director of..." Middle management in large organizations performs the same function as the appendix in the human body. Possibly once useful. Possibly still helpful in some cases. But mostly — just there, and when it becomes inflamed, it can kill.

Smith did not invent the flat structure. The fashionable "flat organization" removes the plates on the doors but leaves the people who sat behind them. Smith invented something more radical: a structure in which every role is justified by function, not by status. On the island there is no deputy for railway-construction strategy. No Chief Survival Officer. There is an engineer who knows how to smelt iron. There is a journalist who knows how to document processes. A hierarchy exists — Smith leads. But it is justified by expertise, not by title.

Compare that with the typical corporation. Between the engineer writing code and the CEO making the product decision — four to eight management layers. Team lead. Engineering manager. Senior engineering manager — a role that exists so that the engineering manager will have someone to report to as well. Director of engineering. VP of engineering — a person whose main job is to synchronize several directors with one another. SVP of engineering. CTO. CEO. Every layer adds delay to decisions and distortion to the transmission of information. Conway's Law¹ does not merely apply — it rages. The organization produces systems that copy its communication structure. Eight layers of management produce eight layers of abstraction in the product.

Smith solves the Conway problem trivially: he shortens the communication path to a single step. Engineer and team. Team and engineer. Idea and execution. No intermediate layers, no "expectation alignment," no "strategy cascading."

Smith's Principle: Four Laws of the Bootstrap Team

Law One: Hire by Competence, Not by Process

The team of The Mysterious Island was not assembled by HR procedures. Cyrus Smith — engineer. Gideon Spilett — journalist. Pencroft — sailor. Herbert — a fifteen-year-old ward of Pencroft with keen natural curiosity. Neb — the universal executor. Each of them made it onto the team not by passing a screening but by possessing skills critical for survival. Verne does not explain the recruitment process further, because there was no process. There was a need for competence and there were people who had it.

The contrast with modern hiring is sobering. In 2026 the path from the first recruiter contact to a start date at a large technology company takes three to six months. Six rounds of interviews. A take-home assignment. Reference checks. A psychometric test that determines whether your anxiety type matches the corporate one. A discussion at the hiring committee. Sign-off on compensation level. Offer letter. Counter-offer. Offer acceptance. A three-month onboarding, during which the new employee learns that half of what he was told at the interview is outdated.

Smith looks at the person. Can the person do what needs to be done today? Does he share the goal of survival? Welcome to the team.

Verne showed this selection in pure form, without a single HR procedure. The team of The Mysterious Island came together not through résumés and not through test scores. Pencroft joined because he could handle any ship. Neb because he was loyal and tough. Fifteen-year-old Herbert — Pencroft's own ward — because he understood the nature of the island better than the adults. Smith asked no one for a diploma or references. He watched one thing: can this person do what needs to be done today, and does he share the goal of survival? Proxy metrics — certificates, service records, test scores — exist in order to lower the risk of hiring the wrong person. Smith's selection is set up the other way around: it maximizes the chance of finding the person who can do the job. On an island where a mistake costs a life, the second approach turns out to be the only one that works. And not by accident, every small team under pressure falls into the same approach the moment the HR department disappears from between the need and the person answering it.

This is not a simplification. It is a different model of selection. Instead of proxy metrics — résumés, certificates, test results — a direct evaluation of the ability to complete the task. Frederick Brooks in The Mythical Man-Month documented that the difference in productivity between the best and the worst programmers reaches a factor of ten.² Not two. Ten. Corporate hiring is optimized for minimizing risk — not hiring the bad person. Smith's hiring is optimized for maximizing the result — finding the person who can.

The difference is fundamental. Minimizing risk scales: you can hire hundreds of "acceptable" people through a standardized process. Maximizing result does not scale: you have to know what you are looking for and be able to recognize it. Smith can put together five people of a perfect team. Not fifty. And in that, his advantage lies.

Law Two: Isolation as a Strategic Decision

The island exists in isolation not by unhappy accident. Verne — and this is important — chooses an uninhabited island not for exotic color but for the purity of the experiment. He removes all external variables in order to show what a team can do when it is not being distracted. Not a single email from headquarters. Not a single call from a stakeholder. Not a single reprioritization because the CEO read an article in the Harvard Business Review. Smith set the goal — build a civilization — and moves toward it without looking over his shoulder at the stock ticker.

In 1943 Clarence "Kelly" Johnson received an order from the U.S. Army Air Forces that would change the history of aviation: build a jet fighter in 180 days. Johnson did something that will look familiar. He took twenty-three engineers and thirty mechanics, moved them into a rented circus tent next to a plastics factory (the smell was so strong that the name "Skunk Works" stuck), and cut them off entirely from Lockheed's corporate bureaucracy.

The result: the XP-80, the first American jet fighter to reach full combat readiness. Delivered in 143 days. Thirty-seven days early.³

Kelly Johnson formulated the fourteen rules of Skunk Works.⁴ The first: "The Skunk Works manager must be delegated practically complete control of his program in all aspects." Smith applied the same principle on the island. Rule number three: "The number of people having any connection with the project must be restricted in an almost vicious manner." Rule number thirteen: "Access by outsiders to the project and its personnel must be strictly controlled by appropriate security measures."

This is not paranoia. It is engineering calculation. Alex Pentland at the MIT Media Lab showed empirically that team productivity is determined not by individual talent but by patterns of communication.⁵ Short, frequent interactions inside the team. Minimal interruptions from outside. Smith's island is an ideal implementation of that pattern — maximum internal connectivity, zero external noise.

Tom Allen at MIT described the effect back in 1977. It was later named the Allen Curve — the probability of regular communication between two people falls exponentially with the physical distance between them. Beyond fifty meters, people communicate no more often than if they were in separate buildings.⁶ Smith's island is a compact base of operations. The whole team sits inside one Allen Curve. Each person is in the zone of spontaneous interaction with every other.

Law Three: The Leader Knows the Subject

Verne constructs Cyrus Smith as the ideal engineer of the nineteenth century — and inside that construction he hides an organizational argument. Smith is not simply "a smart leader." He is a man who calculates the height of a mountain from the angle of a falling stone, synthesizes nitroglycerin from what is at hand, and explains to the team the principle of the telegraph while building it from the metal he himself has smelted from ore.

Verne deliberately makes Smith encyclopedically competent — this is not a weakness of the novel but its thesis. Smith is effective as a leader because he is an engineer. When a technical problem arises on the island, Smith does not call an emergency meeting and does not delegate the decision. He goes to the workbench, to the furnace, to the drawings — and solves the problem personally. Because he knows how everything works. And the team knows that he knows.

This pattern — a leader who owns the subject-matter expertise — repeats in every story of breakthrough products. And every time it repeats, someone writes a business book about it in astonishment.

Steve Jobs was not an engineer in the classical sense, but he understood the product at a level unavailable to the professional managers. Walter Isaacson documents this in the biography — when the Macintosh team showed the prototype, Jobs evaluated not the slides but the product itself.⁷ He opened the case and looked at the layout of the components on the board. He checked how the screen looked in different light. He noticed that the menu font rendered half a pixel off.

From the outside this can be mistaken for micromanagement — both leaders dig into the details. The difference lies in knowledge. The micromanager controls because he does not trust and does not understand. Smith checks because he understands and because he sees what others do not. The micromanager asks "why did you do this?" and waits for apologies. Smith asks "what if we heat it twenty degrees higher?" and waits for the experiment. One controls the process. The other improves the result.

Tracy Kidder described the same pattern in The Soul of a New Machine — Tom West, leader of the MV/8000 project at Data General, was an engineer who understood every aspect of the computer being designed.⁸ His team (the "Microkids" and the "Hardy Boys") worked in a basement, cut off from corporate politics, under a man who could read any diagram on the wall. The project the leadership of Data General had practically written off was completed on time. Kidder's book won the Pulitzer Prize.

Law Four: Crisis as the Normal Mode

Verne is a great screenwriter of catastrophes. In The Mysterious Island he throws at Smith's team a series of existential challenges with mounting intensity — food shortages, pirates, a volcanic eruption, the need to build a whole technical civilization from scratch. These are not literary decorations. They are stress tests of the organizational model. And every time, Smith passes the test the same way. He assesses the situation personally, makes the decision, the team executes. No meetings. No approvals. No escalation.

When the eruption of the volcano approaches and the island must be evacuated, Smith does not convene a "crisis committee" and does not write a "risk management plan." He calculates. How much time until the eruption. What can be saved. What length the hull of a ship needs to be. Arithmetic, not bureaucracy. Engineering calculation, not PowerPoint.

Stanley McChrystal, the general who commanded the Joint Special Operations Command in Iraq, ran into the same problem — the traditional military hierarchy was too slow to fight a networked enemy. His solution, described in Team of Teams, essentially reproduces Smith's model: small autonomous teams that are trusted to make decisions on the spot, without escalation up the chain of command.⁹

McChrystal calls the phenomenon shared consciousness across all participants. Smith arrives at the same result naturally. His team lives in one place, sees the same problems, shares the same fate. When the volcano begins to erupt, shared consciousness arises automatically. An existential threat is a better mechanism for priority alignment than any OKR.

Island Against Cruise Liner

Island: small team, engineer-leader, isolated environment, autonomous decision-making, survival depends on the competence of each member. A mistake by one is a threat to all. Therefore selection by capability, and procedures kept to a minimum.

Cruise liner: thousands of people, dozens of hierarchy levels, passengers do not know how the ship works, decisions are made far from the scene of the action. A mistake by one is an inconvenience to the rest. Therefore procedures replace competence, and regulation replaces common sense.

Most corporations are cruise liners that dream of becoming island teams. They hire thousands of people, create dozens of management levels, build procedures for every eventuality. Then they hire McKinsey consultants at millions of dollars to explain why a five-person startup in a coworking space builds faster.

Melville described this paralysis in 1853, setting it in an office on Wall Street. The scrivener Bartleby answers every request with a single formula — "I would prefer not to" — and this polite formula halts the work of the whole bureau more reliably than any sabotage. The boss can neither compel, nor fire, nor go around him, and gradually builds his own working day around an employee who does nothing. The corporate cruise liner is infected with the same logic, only multiplied across a dozen coordinating instances. A simple change to an interface passes through the designer, the product manager, the architect, the analyst, the release manager — and at every floor sounds a soft "I would prefer to align first." The "Center of Architectural Excellence," without whose visa not a line of code can be touched, is Bartleby's office cast into an org chart — a procedure that exists so that nothing moves without its permission. While the liner is collecting signatures, an island of eight people has already redesigned the interface three times.

McKinsey's answer takes 200 slides and formulates the obvious with such precision that it becomes meaningless.

The problem is not the size. The problem is the architecture. A cruise liner will not become an island no matter how many times you rename departments into "tribes" and "squads." Spotify can afford to call teams squads — they were built as an island. When squads are installed in a bank with fifty thousand employees, you get a cruise liner with pirate flags hung on board, but all the pirates still ask permission before raising the anchor.

Skunk Works: An Industrial Island

The parallel between Smith's team and Skunk Works is not accidental. It is structural.

When Ben Rich (Kelly Johnson's successor) described the principles of Skunk Works, he was essentially retelling the layout of Smith's island — small team, direct access to the engineer-leader, isolation from corporate bureaucracy, hiring by competence.

The results of Skunk Works across half a century: the U-2 (high-altitude reconnaissance), the SR-71 Blackbird (still the fastest air-breathing manned aircraft in history), the F-117 Nighthawk (the first serial stealth aircraft). Each project a breakthrough. Each created by a team smaller than the typical marketing department of a modern IT company.

And now a counterexample. The F-35 Joint Strike Fighter. The most expensive weapons project in the history of humankind. By the GAO's 2024 estimate, the total lifecycle cost of the program will exceed two trillion dollars.¹⁰ Thousands of contractors. Hundreds of thousands of jobs in forty-five states. A political project as much as an engineering one.

The F-35 is a cruise liner. Dozens of levels of decision-making. Every technical decision passes through committees where representatives of the three armed services sit, each with its own requirements. The result, documented in the hearings of the Senate Armed Services Committee: chronic delays, cost overruns, and technical problems that go unresolved for years.¹¹

Kelly Johnson created the SR-71 with a team of a few dozen. The F-35 cannot reach full parameters with a team of tens of thousands. Scaling people does not scale results. It scales coordination, which devours the resources intended for the work.

Brooks formulated this in 1975: "Adding manpower to a late software project makes it later."¹² Smith formulated it in 1875 — gathered five people on an island and built a civilization.

The Dark Side of the Island Team

It would be intellectually dishonest to idealize Smith's model. It has serious limitations, and Verne — to his credit — shows them.

The Problem of Dependence on the Leader

Smith is irreplaceable. If he dies, the island team turns from an engineering civilization into a group of survivors. This is the classic key-person risk — the more competent the engineer-leader, the more fragile the system becomes without him.

In the finale of The Mysterious Island the island explodes and the team is evacuated. But imagine Smith dying in the middle of the novel. The knowledge of how to smelt iron, how to make glass, how to build machines — all of it dies with him. A prophetic metaphor for every startup that has lost its tech lead.

Amy Edmondson of Harvard Business School has researched this phenomenon through the lens of psychological safety.¹³ Teams dependent on a single leader work superbly in the leader's presence and catastrophically in his absence. Resilient teams are the ones where safety is built into the culture, not tied to the personality.

The Problem of Scaling

The island is five people. Not fifty. Not five hundred. Smith's model does not scale linearly, and the attempt to scale it linearly is a typical corporate mistake.

Netflix tried to scale the "island team." Reed Hastings describes in No Rules Rules how the company built a culture of "freedom and responsibility" — minimum rules, maximum autonomy, hiring only A-players.¹⁴ It worked at a hundred employees. At a thousand, it was harder. At ten thousand, Netflix is still more effective than most of its competitors, but the cultural principles require constant active maintenance. An island team that has grown to the size of an archipelago discovers an unpleasant truth — the principles that worked when everybody knew each other by name require evangelism when half the company is made up of people hired last week.

The Problem of the Ethical Compass

Smith is a rational engineer, but even he is not protected from the trap of isolation. When a small team is cut off from the outside world, its ethical compass depends entirely on the leader. The team follows Smith not because the goal is objectively correct, but because it trusts his expertise.

This is the dark side of bootstrap leadership: an autonomous team with a charismatic leader can build with tremendous efficiency in a catastrophically wrong direction. Elizabeth Holmes's Theranos is a modern "island team" with a false mission. A small group isolated from the outside world — literally: employees signed nondisclosure agreements, and the labs were closed to external auditors. A leader with a vision that turned out to be a hallucination.

In 2018 the SEC charged Holmes with massive fraud — raising more than seven hundred million dollars of investment on the basis of false claims about the technology.²⁰ The structure of Theranos copied every principle of bootstrap leadership: small team, isolation, visionary leader, absence of bureaucracy. The only difference — Cyrus Smith actually knew how to smelt iron.

The difference between Smith and Holmes lies not in the structure but in the competence and in the feedback with reality. Smith sees the result every day. The bridge either stands or it collapses. The tool either works or it does not. Holmes claimed to have built a technology but arranged the system so that no one could check. Bootstrap leadership without real expertise from the leader and without a mechanism to verify the result is not an island team. It is a raft whose captain has forbidden looking at the horizon.

Six Signs Your Organization Is a Cruise Liner

Holmes on Baker Street taught us to observe. Apply his method to organizational diagnosis.

Sign one: a decision passes through more than two levels of approval. On Smith's island the path from idea to action is a single step — the engineer decides, the team executes. If your decision travels from team lead to manager to director to VP to another VP, you are not on an island. You are in the line at passport control.

Sign two: nobody can explain why a particular person is needed. In Smith's team every member performs a function without which the civilization collapses. If your organization has titles that can be removed without visible effect on the product, that is dead weight. Not a useful instrument, but a parasitic overlay. The symptom — a person is engaged in "strategy," "processes," or "cross-functional interaction." In other words, coordinates the work of people who would coordinate perfectly well without him.

Sign three: the project leader cannot draw the architecture on a napkin. Smith could describe every mechanism he built on the island. If your CTO cannot explain how your product works without deferring to "the team that handles it" — you do not have a CTO. You have a Chief Title Officer.

Sign four: a crisis produces meetings, not actions. On the island a crisis produces action. In a corporation a crisis produces an emergency conference call with thirty participants, in which the first twenty minutes are spent adjusting microphones. By the time everyone has joined, found the "unmute" button, and agreed on the next meeting to discuss the plan of action, Smith's island team has already solved the problem and is building on.

Sign five: the hiring process is longer than the first project of the new hire. If hiring takes six months and the first project takes three, then your organization spends more effort on attracting people than on using them.

Sign six: reorganizations happen more often than product releases. Smith's team never reorganized. It was assembled correctly from the start. Every corporate reorganization is an admission that the previous structure was a mistake. If reorganizations have become an annual ritual, you are not optimizing the system — you are rearranging the deck chairs. Only not on the Titanic — the Titanic at least had a destination. You are rearranging the chairs on a ship that is circling in the ocean because the captain has forgotten where he was going.

Principles of Bootstrap Leadership

From the experience of Smith, Skunk Works, the Macintosh team, and the Data General MV/8000, a set of repeating patterns emerges. Not a methodology — because the second you print them on a laminated poster and hang it in the open space, they stop working. A set of observations effective exactly until they are turned into a cargo cult.

Size determines effectiveness. Five people on Smith's island. Twenty-three engineers in the first Skunk Works crew. Around a hundred in the original Macintosh team (Jobs held to the rule of "no more than a hundred"). Jeff Bezos formalized it as the "two-pizza rule."¹⁵ A team that cannot be fed with two pizzas is too large. Smith would have fed his team with one wild goat.

Isolation is not a punishment — it is a privilege. Skunk Works moved into a circus tent not because Lockheed was stingy with office space. Because distance from corporate headquarters equals freedom from corporate processes. Smith found himself on an island not because he hated civilization. Because in civilization there are committees.

The leader is first among equals, not first among subordinates. Smith commands. But he commands as a man who can do everything he asks of the team. This is not democracy. It is meritocracy with a single right of veto. Aristotle would have called it "the rule of the best" — a system in which the right to make the decision belongs to the one who understands the task best. Modern HR would call it "toxic leadership." The results speak for Aristotle.

Crisis is a test, not an anomaly. An organization that treats a crisis as a deviation from the norm builds defenses against shocks — and breaks at the first shock that exceeds those defenses. An organization that treats a crisis as part of the norm is antifragile. Smith's team does not panic when the volcano erupts, because every day of life on the island has been an unspoken rehearsal for exactly that. They panic if the lava will destroy their one ship. The difference — between fear of the unknown and respect for the specific threat.

The Dumas Factor: Mission Above Hierarchy

Smith's island team solves the coordination problem through competence. But what if the challenge requires not merely competent people, but people willing to risk everything for a common goal?

Alexandre Dumas in The Three Musketeers showed a different model of a high-performance team — four men (d'Artagnan, Athos, Porthos, Aramis) who act under the motto "One for all, and all for one." Not a bootstrap team of engineers, but a mission-driven team where personal loyalty and the common goal matter more than formal structure.

The musketeers are officers of the royal guard, but they regularly act contrary to direct orders when those orders contradict their understanding of what is right. This is autonomy grounded in shared values, not rebellion against authority.

Translated to the corporate language: a team that shares a mission makes better decisions than a team that executes instructions.

Musketeers Against Committees

In The Three Musketeers the Queen of France receives a threat — and four men save the dynasty, acting without approval from ministers, cardinals, and palace bureaucracy. More than that: they save the queen from that bureaucracy, which is ready to sacrifice her to political compromise.

Sound familiar? Every product manager who bypassed the "approval procedure for new features" to do what the user needed. Every engineer who fixed a critical bug in production without waiting for the release committee. Every designer who redesigned an interface because he knew that if he waited for feedback from every stakeholder, the product would die first.

The difference between the musketeers and the corporate rebels — legitimacy. The musketeers serve the king, and therefore their disobedience to ministers is justified. The team has a higher mission (defense of the crown) that matters more than the intermediate instructions.

Netflix: Musketeers of the Corporate World

Netflix under Reed Hastings created the closest corporate analog to the musketeers' team. The principle of the "Keeper Test" — each employee stays with the company as long as the manager is willing to "fight for them" if they wanted to leave.¹⁶ High standards, but absolute loyalty to those who meet them. Essentially — the professional version of the musketeers' "one for all": we support each other as long as each deserves the support.

The result: Netflix employees make decisions like the musketeers — from the mission (to entertain the world) rather than from the procedures. While competitors spent months negotiating content licenses, Netflix teams bought series rights over a weekend. Where traditional media companies required detailed plans and presentations, Netflix greenlit projects on the basis of pitches and track record. The queen's musketeers acted without the cardinal's approval; Netflix acted without the media industry board's.

Limitations of the Musketeer Model

Like Smith's island team, the musketeer model has its dark sides.

The problem of scaling. "One for all" works for four people. For forty — already harder. For four thousand — practically impossible, because the personal mission that held the original group together dissolves into the layers of management until it turns into a bullet point on the corporate-values wall of the meeting room. Personal loyalty does not scale as efficiently as procedures.

The problem of selection. The musketeer team requires not merely competence but cultural fit. One cynic can destroy the atmosphere of mutual trust on which the whole model rests. Because of this, hiring into a musketeer team turns out to be an even more selective process than into Smith's island team — here it is not enough merely to possess the required skills.

The problem of direction. The musketeers are loyal to the king. But what if the king is wrong? Mission-driven teams are dangerous when the mission is false or the leader defining the mission loses the compass. Theranos again a good example — a team of "musketeers" loyal to a false vision.

Hybrid Model: Island of Musketeers

The best technology teams in history combined the principles of Smith and Dumas: the technical competence of the island team with the missionary devotion of the musketeers.

The team of the first iPhone — islands of competence (engineering, design, product) under a single command (Jobs), united by a mission (to reinvent the mobile phone) and ready to go against the whole mobile industry.¹⁷

The early team at SpaceX — an island of engineering competence plus the musketeers' devotion to the mission of making humanity multi-planetary.¹⁸

The team at DeepMind — an island of artificial-intelligence researchers with a mission to solve intelligence and, with its hands, to solve everything else.¹⁹ DeepMind is also proof of the inverse of Conway's Law — if you want to create something fundamentally unified (general artificial intelligence), build a team that is fundamentally connected. Not a neural-networks department, a reinforcement-learning department, and an applications department. One research group, one mission, one Allen Curve.

The pattern repeats: a small team of experts, united by an ambitious mission, isolated from the corporate mainstream, under a leader who understands both the technology and the vision.

Smith built a civilization on an island. The musketeers defended a queen against a conspiracy. The best teams in the history of technology do both: create a new reality and defend it from those who would turn innovation into committees.

But Verne, to his credit, does not end The Mysterious Island with triumph. The island explodes. The team loses everything it built. They are saved by Nemo — a deus ex machina they did not deserve and did not control. The most honest metaphor of bootstrap leadership: you can build the perfect team and the perfect system — but a volcanic eruption does not create a ticket in JIRA and does not wait while you conduct a retrospective.

To be continued. In the next chapter: London, 1897 — while the vampire hunters play the hero each on his own, Mina Harker reduces the diaries, letters, and newspaper clippings into a single card index. The count will lose not to the wooden stake — but to the assembled information.

Footnotes:

¹ Conway, Melvin. "How Do Committees Invent?" Datamation, April 1968. Canonical English text used directly. Original formulation of Conway's Law: "Organizations which design systems are constrained to produce designs which are copies of the communication structures of these organizations."

² Brooks, Frederick. The Mythical Man-Month (1975). Addison-Wesley. Canonical English text used directly. Data from Sackman, Erikson, and Grant (1968) on the productivity variance among programmers, ranging from 1:10 to 1:28.

³ Rich, Ben R., and Leo Janos. Skunk Works: A Personal Memoir of My Years at Lockheed (1994). Little, Brown and Company. Canonical English text used directly. Note: the first American jet in flight was the Bell P-59 Airacomet (October 1942), but the P-59 did not reach combat readiness; the XP-80 / P-80 Shooting Star became the first U.S. jet fighter fit for operational use.

⁴ Johnson, Clarence L. "Kelly's 14 Rules & Practices." Lockheed Martin Skunk Works. Originally formulated in 1943, later refined by Ben Rich. Full text reproduced in Rich and Janos (1994). Canonical English quotations used directly.

⁵ Pentland, Alex. Social Physics: How Social Networks Can Make Us Smarter (2014). Penguin Press. Canonical English text used directly. Studies from the MIT Human Dynamics Lab showed that communication patterns explain up to 35 percent of the variance in team productivity (see also "The New Science of Building Great Teams." Harvard Business Review, April 2012).

⁶ Allen, Thomas J. Managing the Flow of Technology (1977). MIT Press. Canonical English text used directly. The "Allen Curve" — exponential decrease in communication frequency with distance.

⁷ Isaacson, Walter. Steve Jobs (2011). Simon & Schuster. Canonical English text used directly. Description of the work on the first Macintosh, including the requirements for the aesthetics of internal components.

⁸ Kidder, Tracy. The Soul of a New Machine (1981). Little, Brown and Company. Canonical English text used directly. Pulitzer Prize for General Nonfiction, 1982.

⁹ McChrystal, Stanley. Team of Teams: New Rules of Engagement for a Complex World (2015). Portfolio / Penguin. Canonical English text used directly.

¹⁰ U.S. Government Accountability Office. F-35 Joint Strike Fighter: Cost Growth and Schedule Delays Continue. GAO-24-106909, May 2024. Canonical English text used directly. Updated estimate of the total lifecycle cost of the program — above $2 trillion, versus earlier estimate of $1.1–$1.7 trillion (GAO-21-105282, April 2021).

¹¹ U.S. Senate Committee on Armed Services. Hearings on the F-35 program, held repeatedly across 2014–2021. Canonical English text used directly. Documented problems include software-development delays, issues with the ALIS / ODIN system, and cost overruns.

¹² Brooks, Frederick. The Mythical Man-Month (1975). Addison-Wesley. Canonical English quotation: "Adding manpower to a late software project makes it later" — Brooks's Law.

¹³ Edmondson, Amy. The Fearless Organization (2018). Wiley. Canonical English text used directly. Research on the connection between psychological safety and team effectiveness.

¹⁴ Hastings, Reed, and Erin Meyer. No Rules Rules: Netflix and the Culture of Reinvention (2020). Penguin Press. Canonical English text used directly.

¹⁵ Bryar, Colin, and Bill Carr. Working Backwards: Insights, Stories, and Secrets from Inside Amazon (2021). St. Martin's Press. Canonical English text used directly. The "two-pizza rule" is described based on Amazon internal practices from the mid-2000s onward.

¹⁶ Hastings, Reed, and Erin Meyer. No Rules Rules: Netflix and the Culture of Reinvention (2020). Penguin Press. Canonical English text used directly. Description of the Keeper Test and the high-performance culture.

¹⁷ Young, Jeffrey S., and William L. Simon. iCon Steve Jobs: The Greatest Second Act in the History of Business (2005). Wiley. Canonical English text used directly. Documentation of the culture of Apple teams under Jobs.

¹⁸ Vance, Ashlee. Elon Musk: Tesla, SpaceX, and the Quest for a Fantastic Future (2015). Ecco. Canonical English text used directly. Description of the formation and culture of the early SpaceX teams.

¹⁹ Suleyman, Mustafa. The Coming Wave (2023). Bodley Head. Canonical English text used directly. Insider perspective on the creation and evolution of the DeepMind research culture.

²⁰ U.S. Securities and Exchange Commission. Litigation Release No. 24065, SEC v. Elizabeth Holmes and Theranos, Inc. March 14, 2018. Canonical English text used directly.