Antifragility by Craft: Escaping Iatrogenics in Engineering

The Seduction of the Fix

We live in a culture obsessed with optimisation. Every defect must be resolved. Every shortcoming gets a hotfix. Entire teams exist to support the act of fixing. But what if the very act of relentless correction is part of the problem?

The modern tech stack is a patchwork. Behind the language of iteration, sprint velocity, and feedback loops, there is often a deeper truth: we are building too fast, and too shallowly. Instead of designing for permanence, we treat stability as a future concern, a phase we reach after the chaos.

Nassim Nicholas Taleb calls it iatrogenics: harm caused by the healer. A term borrowed from medicine, it applies just as well to organisations and systems. When we overcorrect, we destabilise. When we touch things constantly, we stop trusting them. We fall into the trap of overengineering, layering complexity over complexity, without a unifying concept or consistent foundation.

This breaks a fundamental principle of engineering: complex systems fail more. The fix becomes the disease.

Craftsmanship Does Not Wait for Approval

True craftsmanship begins without permission. It does not rely on a process, a task, or a stakeholder to authorise it. The craftsman does not wait to be told to care. He simply does.

This mindset has existed across centuries: shipbuilders, clockmakers, bridge architects, gunsmiths. These people worked in constraints far harsher than ours. And yet they produced systems that endure, not because someone asked them to, but because they held the work to a higher standard. And in doing so, they laid the groundwork for much of the modern world. The systems we depend on, from transit to telecommunications, stem from their principles, precision, and refusal to compromise.

In software, we have diluted the concept of quality into "clean code" or "tech debt management." But this is a shadow of what real craftsmanship meant. Craftsmanship is not a checklist. It is an ethic.

To the craftsman, speed does not matter if it robs clarity. Approval does not matter if it replaces integrity. One does not require permission to build well, only the courage to care when no one is watching.

Build It Once. Build It Well. Fix It Never.

The best systems are not the ones we fix fast. They are the ones we rarely need to fix at all. And that simple truth tears apart the hero culture, the one that celebrates midnight saves and patch warriors instead of quiet, enduring design.

To build well is to front-load responsibility. It is to think about the lifetime of the thing you are building, not just the feature you are shipping. It is to say: "I will not pass this problem forward."

There is a difference between agility and fragility. The former adapts. The latter collapses. But too many engineering cultures blur this line, confusing constant change with resilience.

The truth stands simpler: if you build it right, it resists chaos. It absorbs stress. It grows slowly but confidently. Just like the old rifles that still fire. Just like stone bridges still standing centuries later. Just like UNIX tools doing one thing well.

To fix less is not to do less. It is to build deeper, with fewer assumptions, more judgement, and longer-term orientation.

The Hidden Cost of Over-Fixing – Iatrogenics

The danger of over-fixing is not just wasted effort. It is erosion of trust and increase of entropy.

Taleb’s concept of iatrogenics warns us: some interventions do more harm than good. In medicine, these are treatments that create side effects worse than the original disease. In engineering, these are patches that introduce regressions, workarounds that obscure intent, and optimisations that kill simplicity. Caring does not mean overthinking or over-medicating. It means expecting the unexpected while designing for it.

Because Taleb was right: centralised systems, rigidly planned to handle everything, often collapse under pressure. The craftsman builds differently, with optionality, redundancy, and margin. The result is not a perfect system, but an antifragile one: a system that endures attacks, adapts through failure, and quietly grows stronger.

In Taleb’s terms, the craftsman avoids Extremistan, which means the world of rare, catastrophic events. It avoids it by deliberately building for Mediocristan, where variation is mild and outcomes are predictable.

Extremistan is where a single failure can bring everything down, like a security breach that compromises millions, or a service dependency that collapses the whole stack. In contrast, Mediocristan keeps errors local and manageable, a world where damage does not scale catastrophically.

Every unnecessary fix adds surface area. It amplifies risk. It detaches the builders from the system’s essence. And worse, it invites the kind of systemic fragility Taleb warns us about, the kind that pulls us out of Mediocristan and into Extremistan, where rare but catastrophic events dominate. The Black Swans we thought improbable, or outright denied, suddenly appear, and what happens next is always worse than we imagined.

The Return of the Quiet Professional

Not everything requires noise. Not everything must move fast. Sometimes the highest form of contribution is silence, because the system simply works.

There is a place in every field for the quiet professional. The one who builds well, tests thoroughly, documents clearly, and leaves no broken glass behind. The craftsman’s mindset shapes more than the work. It shapes the way they behave, the process they follow, and the systems they leave behind. These systems mirror their character: discreet, resilient, and humble. And the beauty of it? That mindset ripples outward, ultimately reaching the end customer as clarity, reliability, and trust.

To build well is to shield the organisation from future harm. To reduce the need for heroics. To leave space for invention instead of interruption.

The craftsman leaves behind fewer bugs, fewer meetings, fewer surprises. Not through perfection, but through care and foresight.

What We Build Reflects What We Believe

Fixing remains easy. Building right remains hard. And in many organisations, the easy thing has become the default.

Craftsmanship requires no permission.

It demands a deliberate choice: to design with longevity in mind, to resist the noise, and to accept that, on occasion, the most effective fix lies in constructing something that endures without intervention.

This is not a gesture but a quiet discipline, one that begins with the individual, shapes the process, persists within the system, and ultimately manifests in the experience of those who rely upon it.