AMAtomic Modules
Atomic Modules Chapter 16Menu
Part IV · Reusing & Swapping

Swapping at Any Altitude

The third move, swapping: change one piece while the larger whole keeps working.

The Move That Makes Work Durable

You can combine and you can repackage. The third move is swapping, and it is the one that decides whether your work survives the world changing around it or gets destroyed by it. Everything you build eventually faces change. A number goes stale. A law is rewritten. An example ages out. A whole approach is overtaken by the field. For most people, change means tearing down and rebuilding, because their work is a monolith and you cannot edit a monolith without remaking it. For you, change means something far smaller. It means a swap.

Swapping is replacing one piece while the larger whole keeps working. You take out the part that no longer serves and put in one that does, and everything around it holds, because the part had clean edges and the whole never depended on that part’s internals. This is the move that turns change from a catastrophe into a maintenance task, and it is the final piece of the case this book has been building: that work made of clean modules is not just faster to create, it is durable in a way monolithic work can never be.

Atomic idea

Change becomes a swap, not a teardown. You always still have the course.

Definition

Swap Replacing one piece with another while the larger whole keeps working, done at whatever altitude the change actually happened.

Replace the Piece, Keep the Whole

The principle is simple and worth stating plainly. You replace a piece at its own level, and the level above survives untouched.

Why does the whole survive? Because of everything you built in the earlier parts. A clean module connects to its surroundings through its edges, not its insides. The course relies on the compound-interest lesson to do its job; it does not rely on the particular example inside that lesson. So you can change the example, the inside, freely, and the course never notices, because what the course depended on, the job, is still being done. Swapping works precisely to the degree that your boundaries are clean. A piece with tangled edges cannot be pulled without tearing what it was fused to. A piece with clean edges lifts right out, and a replacement drops right in. Every bit of discipline you spent on self-containment in Parts II and III was, in part, an investment in this moment, the moment you need to change one thing without disturbing everything.

Any Altitude

The reason swapping deserves its own chapter is that it works at every level, and the level you swap at depends only on how much has changed.

Sometimes what changed is small, a single chunk. In your finance course, a contribution limit you cite goes up for the new year. That figure is one chunk inside the investing lesson. You swap the chunk: out with the old number, in with the new one. The lesson holds, because the lesson was never about that exact figure; it was about the principle, and the figure was just one replaceable part of it. A two-minute fix, and the course is current again.

Key idea

Swap at the altitude where the change actually happened, and only that piece moves.

Sometimes what changed is larger, a whole module. Suppose the field moves and an entire lesson is now built on an approach nobody recommends anymore. This is not a figure to update; the whole lesson has to go. So you swap at the module level: you pull the outdated lesson out of the course and drop a new one into the slot it occupied. The course survives the substitution, because the lesson was a clean, self-contained part with clear edges. You did not rebuild the course. You replaced one of its components. And the same logic runs all the way up: an entire course inside a program can be swapped for a better one, and the program holds. Swap at the altitude where the change actually happened, and only that piece moves.

You Still Have the Course

There is a feeling that comes with this, and it is worth naming because it changes your relationship to change itself. When something you teach becomes outdated, the old dread is that your work is ruined and you are back to square one. The truth, once your work is modular, is the opposite. You still have the course.

Say that outdated lesson scared you. Without modules, an outdated lesson can feel like an outdated course, because everything is welded together and you cannot see how to fix one part without redoing the whole. With modules, the damage is contained to exactly the piece that changed. The lesson is outdated; the course is not. You swap the one lesson and everything else you built, the structure, the other lessons, the sequence, the years of refinement, is all still there, still yours, still working. Change stops being a threat to the whole and becomes a localized repair. That is an enormous shift in how it feels to maintain a body of work over years: not a series of teardowns, but a series of small, confident substitutions.

Worked: A Change, Two Altitudes

Put both levels in one pass. It is January, and two things have changed in your finance course.

First, that contribution limit went up. The old way, if the figure were woven through a monolithic lesson, you would reread the whole thing hunting for every place the number appears and every claim that depends on it, hoping you caught them all. The swap way, the figure is a chunk, ideally a chunk that lives in one place as a single source of truth, so you change it once and the lesson is correct. Minutes, not an afternoon, and no risk of a stale copy hiding in a corner. The single source of truth told you where to change; the swap told you how much to touch.

Second, a lesson on a particular savings vehicle is now obsolete; the rules around it changed so much that the lesson teaches the wrong thing. This is a module-level swap. You build a clean replacement lesson, pull the obsolete one out of its slot, drop the new one in, and write one new bridge at the seam. The course is whole again. Notice what you did not touch: every other lesson, the order, the opening, the close. One module changed; the rest stood. Same course, brought current at two different altitudes, with two different sizes of swap, and at no point did you face rebuilding the thing. A swap is not finished when the new piece drops in; it is finished when the seam works and the whole still flows.

Clean Edges Everywhere

The trades show the same move so consistently that it is almost the definition of well-built work. A car maker issues a recall and swaps one faulty component across an entire product line without rebuilding the cars, because the component had a defined interface. An understudy steps into a role and the show goes on, because the production was built so a part could be replaced. And you have already lived this yourself: when Atomic Chunks was edited, overused example chunks were swapped out for fresh ones without rewriting the chapters around them, because each example was a replaceable part, not load-bearing structure. Whole chapters could have been swapped the same way. In every case, the swap was painless for exactly one reason, the boundaries were clean. That is the through-line of this entire book arriving at its sharpest point: build in clean, self-contained pieces, and you buy yourself the ability to change anything, at any level, without remaking everything.

Swap at the Right Altitude, and No Higher

One skill keeps swapping efficient: matching the swap to the size of the change, and resisting the urge to go higher than you need to. The question to ask when something changes is narrow on purpose. What, exactly, stopped being right? Then swap at that level and no higher.

The common error is over-swapping, rebuilding a whole lesson because one figure in it went stale, or rewriting a whole course because one lesson aged out. That is mistaking the altitude of the change. If only a number changed, the change lives at the chunk level, so the fix lives there too; touching the whole module is wasted work and risks introducing new mistakes into parts that were fine. The opposite error, under-swapping, is patching a single chunk when the entire approach is obsolete, leaving a fundamentally outdated lesson wearing one fresh number. Diagnose honestly: locate the lowest level at which the thing is actually wrong, and swap exactly there. The cleaner your boundaries, the more precisely you can target, which is one more reason the discipline pays off. Good modular work lets you make the smallest correct change, and the smallest correct change is almost always the right one.

What Comes Next

Swapping completes the three moves. You can replace a piece at its own altitude, a chunk or a whole module or a whole course, and the level above survives, because clean boundaries let parts come out and go in without tearing the whole. Your work becomes durable: change is a localized swap, not a teardown, and you always still have the course.

But swapping assumes you have something to put in the empty slot. Often you will not. A project needs a module you have never built, or a slot opens that nothing in your library fills. That is the last situation Part IV has to handle: what to do when there is a gap. You can build, borrow, swap in something close, or cut, and whatever you build to fill the gap, you bank, so the gap never has to be filled twice. That is the next chapter.

Try this

Diagnose the altitude of a change.

Think of something in your work that recently went out of date. Ask exactly what stopped being right, a single figure, a whole lesson, an entire approach, and swap only at that level. The smallest correct change is almost always the right one.