Notes

• 30 October 2026

  A design can be technically correct and still fail its purpose. Sometimes, defending the principles becomes easier than questioning whether they still serve the problem.

• 10 October 2026

  An organisation can become remarkably good at working around a problem without ever becoming good at solving it. Eventually, the workaround becomes part of the process, and the original problem stops looking like a problem at all.

• 8 October 2026

  Systems rarely fail on the day we notice them. By then, people have usually spent months adapting to things that should have been corrected. The interesting question is not why the system finally broke, but when everyone started treating what was wrong as normal.

• 7 October 2026

  Some ideas take years to develop and only a few minutes to explain. This creates an odd imbalance: the clearer the explanation becomes, the easier it is to underestimate the work behind it.

• 5 October 2026

  The same event can be described as a success, a compromise, or a warning, depending on the time horizon used to examine it. Short-term results have a habit of borrowing certainty from the future.

• 2 October 2026

  We often assume that more information will improve a decision. Beyond a certain point, additional information can become another way of postponing responsibility for choosing.

• 30 September 2026

  Publishing a thought changes one's relationship with it. Until then, it can be revised indefinitely. Afterwards, it begins to encounter interpretations that were never part of the original intention.

• 23 September 2026

  An unfinished idea is not necessarily an incomplete idea. Sometimes it has reached the limit of what can be understood without allowing the world to respond.

• 17 September 2026

  The usefulness of a model depends partly on what it excludes. A model that attempts to represent everything eventually becomes another version of the problem it was meant to simplify.

• 8 September 2026

  There is a difference between a complicated system and a complex one. A complicated system may require considerable knowledge to understand. A complex system can behave unexpectedly even after its individual parts are understood. Confusing the two leads to considerable overconfidence in planning.

• 3 September 2026

  Writing is a peculiar form of thinking. An idea can remain perfectly convincing inside the mind until it is forced into a sentence. The sentence exposes what the thought had been allowed to conceal.

• 28 August 2026

  We tend to think of freedom as having more choices. But every additional choice also creates another possibility to evaluate, reject, or regret. Sometimes freedom is simply having fewer things that require a decision.

• 19 August 2026

  A new beginning is rarely as clean as the phrase suggests. Old responsibilities, habits, and ways of thinking continue to influence decisions long after the circumstances that produced them have changed.

• 6 August 2026

  Distance changes the apparent size of problems. Things that once occupied an entire day sometimes become surprisingly small when they no longer demand an immediate response.

• 30 July 2026

  Leaving a system does not immediately remove it from the mind. For some time, one continues anticipating its problems, remembering its dependencies, and mentally completing conversations that no longer require a response.

• 16 July 2026

  One of the stranger things about professional life is how easily a person's contribution becomes part of the background. The work remains visible, the effort gradually disappears from the story.

• 2 July 2026

  We often confuse familiarity with belonging. A place can become familiar because we have learned its constraints, routines, and exceptions. That does not necessarily mean we belong there.

• 25 June 2026

  A difficult decision often becomes easier once the question changes from what can be preserved to what is worth preserving. These two are not always the same.

• 12 June 2026

  Some organisations measure productivity by how quickly people respond. This creates a peculiar environment in which uninterrupted thinking appears less productive than repeatedly acknowledging interruptions.

• 7 June 2026

  There comes a point when knowing how to solve a problem is no longer sufficient reason to take ownership of it.

• 26 May 2026

  A good architect is not necessarily the person who knows the most technologies. It may be the person who knows which decisions should remain reversible, and for how long.

• 14 May 2026

  The word alignment is used rather generously in organisations. Often it means that everyone has heard the same presentation, not that everyone has understood the same problem.

• 3 May 2026

  Not every delay is indecision. Sometimes the mind has already rejected the available choices but has not yet found a defensible alternative.

• 24 April 2026

  A system becomes difficult to change when too many people have learned to work around its limitations. At that point, the limitations are no longer merely technical. They have become part of the organisation's operating knowledge.

• 11 April 2026

  The most expensive technical decisions are not always the ones involving the largest investments. Sometimes they are small assumptions made early, embedded deeply, and discovered too late to reverse economically.

• 2 April 2026

  Sometimes a meeting ends with everyone agreeing because the difficult question was never asked. Agreement is easy when the consequences are still abstract.

• 29 March 2026

  Experience is not simply the accumulation of things that worked. Much of its value comes from remembering what almost failed, what required intervention, and what should never have been attempted in the first place.

• 17 March 2026

  Technology does not remove organisational ambiguity. It usually gives ambiguity a faster and more consistent way to travel.

• 9 March 2026

  There are people who ask questions because they do not understand, and people who ask questions because they understand enough to be uncomfortable. The second kind is frequently mistaken for resistance.

• 4 March 2026

  A dashboard can tell you that everything is green. It cannot tell you how much effort was required to keep it that way. The colour describes an outcome, not the condition of the system producing it.

• 27 February 2026

  There is a difference between understanding a problem and being able to explain it. The first can happen in seconds. The second may take hours, particularly when the explanation must survive people who were not present when the problem emerged.

• 21 February 2026

  The more experienced someone becomes, the less frequently they should need to demonstrate their experience. Yet many environments reward exactly the opposite.

• 18 February 2026

  A decision can be correct when made and wrong when examined later. The difficulty is that organisations often evaluate yesterday's judgement using information that became available only today. This makes hindsight a remarkably unfair performance metric.

• 15 February 2026

  Most leadership metaphors are wrong. They glorify control. Real systems reward navigation under partial control.

• 14 February 2026

  Success depends on system timing as much as capability. Careers stall less from lack of talent and more from lack of timing.

• 13 February 2026

  Corporate is closer to Ludo than chess. Not just strategy, but timing + placement + chance.

  You don't control the dice, you control what you do after it rolls.

• 13 February 2026

  Over time, I learned that reacting immediately was not the problem, sending immediately was. I would write the reply at full speed, with everything I felt in it, knowing it could not be sent. The emotion needed somewhere to go. Once it was out, the body settled. The anger remained, but it became quieter, usable. Only then did editing begin. Logic entered later, not first.

  Restraint was not silence. It was sequencing.

• 12 February 2026

  While my flight was still moving after landing, many passengers immediately switched on their phones. What was visible was not impatience, but conditioning.

  Systems train people to act on signals rather than judgment and over time, behavior continues even when purpose disappears. Urgency survives long after necessity and habit replaces understanding.

• 11 February 2026

  The human brain is not built for parallel processing. It is built for sequencing.

  What often looks like multitasking is usually rapid context switching, and even that works only for a few. Most people who appear to handle many things at once are actually running a tight internal queue.

  The efficiency comes not from doing everything together, but from understanding everything early, retaining it mentally, and then processing items one by one based on priority.

  This demands two things that are rarely discussed together: high retention without clutter, and the ability to switch contexts without residue. Without both, parallelism degrades into noise. With both, sequencing can outperform any illusion of multitasking.

• 9 February 2026

  Sometimes the most effective response to ambiguity is not resistance, but containment. When a requirement refuses to specify its own boundaries, the only leverage left is to accept it exactly as written, and no more.

• 8 February 2026

  Long ago, I stopped treating memory as storage and started treating it as bandwidth. The more trivial details I offloaded, the more space there was for patterns, relationships, and judgements. Forgetting became a feature, not a flaw. How much you remember matters less than what your mind stays free to notice.

• 28 January 2026

  More often than not, systems function only because someone temporarily collapses multiple roles into one. Once the moment passes, the structure remains unchanged.

• 13 January 2026

  Career grows upward, operating style grows inward.

• 12 January 2026

  The first boss matters not because of authority, but because they normalise a way of working that feels "successful".

• 11 January 2026

  If chaos, shouting, and pressure produced results once, the pattern hardens. Seniority amplifies it; it does not correct it.

  If quiet competence, hands-on problem solving, and grounded decision-making worked early, the same person often builds teams that trust rather than fear.

• 10 January 2026

  Operating style is usually formed early, not refined later. What works in the first few years becomes default bahaviour.

• 6 January 2026

  Some teams are not behind schedule. They are running on borrowed judgement. The invoice is never itemised.

• 12 December 2025

  Most work is not parallel. It is queued, held, and sequenced internally. What looks like multitasking is usually fast switching plus retention.

• 2 November 2025

  Confidence built on acceptable outcomes is dangerous. It teaches systems that effort is infinite and judgment is optional.

• 9 August 2025

  Processes are very good at winning arguments. They are less good at noticing when work had to move around them to succeed.

• 27 May 2025

  Silence is often mistaken for disengagement. In reality, it is frequently the space where complexity is being held together.

• 15 March 2025

  Scope is not argued; it is documented. Responsibility is not distributed; it is made explicit. What the system refuses to clarify, it must carry itself.

• 14 February 2025

  Stability created by personal sacrifice is not stability. It is debt that compounds silently.

• 19 November 2024

  Roles describe where responsibility should sit. They say very little about where work actually appears. This gap is where most systems leak energy.

• 3 September 2024

  Some costs never appear on balance sheets. They appear as hesitation, fatigue, and people quietly lowering their ambition.

• 11 June 2024

  In many organisations, continuity is mistaken for correctness. If work keeps moving, design is assumed to be sound. This is how fragile systems gain confidence.

• 22 March 2024

  When nothing breaks, the system congratulates itself. It does not ask who bent, stretched, or stayed late to make that possible.

• 14 February 2024

  Most systems don't fail because of bad intent. They fail because responsibility travels faster than authority. When someone absorbs that gap quietly, the system survives. When no one does, system escalates. Escalation is not a solution. It is a symptom that absorption has run out.

• 28 January 2024

  People confuse being visible with being reliable. Visibility creates memory while reliability creates responsibility. Systems reward the first loudly and they rely on the second silently.

• 8 January 2024

  Many organisations survive not because they are well-designed, but because a few people repeatedly absorb what the design cannot handle.

• 19 May 2023

  People who keep systems running quietly are rarely rewarded. People who announce problems loudly often are. This explains more behaviour than most incentive frameworks.

• 5 December 2022

  The best systems don't eliminate discretion. They place it where the consequences are closest.

• 22 August 2022

  Systems rarely collapse because of failure. They collapse when no one is left who can compensate.

• 30 July 2022

  Authority decides who may speak. Judgement decides what should be said. They are rarely aligned.

• 7 April 2022

  The most stable performers are not the fastest. They are the ones who can hold multiple incomplete thoughts without losing clarity.

• 11 February 2022

  If escalation becomes a habit, it stops being a signal. It becomes background noise; and real emergencies are ignored.

• 27 September 2021

  Speed without judgement creates urgency. Judgement without speed creates irrelevance. Most organisations oscillate between the two and call it agility.

• 9 April 2021

  Good professionals reduce noise before it becomes visible. That is why they are often accused of "not doing much". The absence of chaos is rarely credited to those who prevented it.

• 15 March 2021

  Habit can outlive necessity. This is how unnecessary processes survive.

• 18 October 2020

  The most dangerous phrase in large organisations is not "This is impossible", but "This is not my responsibility."

• 28 June 2020

  When people stop questioning processes, they are no longer following structure. They are following memory.

• 2 June 2020

  Meetings don't waste time individually. They waste it collectively. The cost is not one hour; it is one hour multiplied by the number of people who didn't need to be there.

• 22 January 2020

  Systems rarely collapse where they are weakest. They collapse where people assume someone else is watching.

• 8 December 2019

  Systems train people to act on signals rather than judgement. Urgency persists even after purpose disappears.

• 3 December 2019

  Most organisational problems don't come from people doing wrong things. They come from people doing right things in a structure that rewards the wrong outcomes. Over time, the structure wins. Not because it's smarter, but because it's persistent.

• 7 August 2019

  When someone says, "This is how the process works," what they often mean is, "I don't want to think about this right now." Process is useful, abdication dressed as process is not.