The Answer is 42
On causal and enumerative thinking
In The Hitchhiker’s Guide to the Galaxy, a civilization builds a supercomputer to answer the ultimate question — life, the universe, and everything. It runs for seven and a half million years and returns a single number. Forty-two. The answer is correct. The trouble is that nobody had worked out the question.
You do not need a supercomputer to make this mistake. Organizations make it by hand, every day, and lately at a speed that should worry us.
There are two ways organizations think.
The first is causal. It begins with a model of the world. Something is changing. That change creates new constraints. Those constraints reveal a point of leverage. From that leverage, the work follows. This kind of thinking has a thesis, a theory of value, a proposed path, and a way to be wrong. If reality disagrees, the model breaks. That is a feature, not a bug.
The second is enumerative. It begins with a list. Projects, owners, timelines, workstreams, diagrams, integrations, stages, milestones. Everything is arranged. Everything has a name. Everything appears coordinated. But nothing is derived.
Both can look like strategy. Only one is. The causal kind turns uncertainty into a model. The enumerative kind turns ambiguity into a portfolio.
I have been handed both kinds of document. The enumerative one is almost always the more impressive object. It is longer. It has more boxes, more owners, more arrows. It looks like more work went into it. And it tells you the least about whether any of the work is worth doing. I did not always know to be suspicious of the better-looking document. I learned it the slow way.
The cost of a missing model
A causal document reduces future cognitive work. Once the model is written, people can derive the work from it. What does it imply? What follows from the thesis? What would contradict it? What should we build first? What should we not build at all?
An enumerative document increases future cognitive work. Because the model is missing, every future question requires interpretation. What did we mean by this? Does it fit? Who owns it? Is it in scope? How does it connect? What do we do next?
Causal thinking front-loads the thinking. Enumerative thinking institutionalizes repeated interpretation.
This is why the causal kind feels liberating to builders. It lets them solve a problem, commit to a direction, ship something, encounter reality, and move on to the next problem. The thinking has been embedded into the model. The work continues without the author hovering over it.
That is what good engineering does. A good abstraction makes the person who wrote it less necessary. A good API removes repeated explanation. A good test suite removes repeated anxiety. A good architecture removes repeated debate. A good strategy should do the same. The goal is not to avoid thinking. The goal is to avoid thinking twice about the same unresolved question.
The document that cannot be wrong
A causal strategy can fail. If it says a certain capability will create a certain outcome, and the outcome does not appear, the model must be revised. Reality has a way to respond.
An enumerative strategy rarely fails as a strategy. Individual projects fail. Timelines slip. Teams change. But the umbrella survives, because it never made a sharp enough claim. Anything that works gets folded underneath it. Anything that fails is explained away as incomplete execution, shifting priorities, or the need for another phase. It does not meet reality directly. It keeps moving sideways.
That difference shapes who holds the power.
In causal thinking, authority follows the model. The person who defines the causal chain has embedded the future reasoning into the artifact. If the model is good, others execute without constant interpretation. Authority migrates from the person into the document.
In enumerative thinking, authority follows the interpreter. Because the artifact is under-specified, someone must always explain what it means. Every new question, every new project, every new buzzword, every new ambiguity returns to whoever controls the vocabulary. If the model is never fully stated, the interpreter never becomes obsolete. There are always clarifications to make, approvals to grant, meanings to authorize.
The thing I find hardest to say plainly is that the missing model is not always an accident. A document that has to be interpreted keeps its interpreter necessary. I have watched ambiguity defended as flexibility when what it really preserved was someone’s place in the chain. I have probably done a smaller version of it myself, without admitting that was what I was doing.
Causal models make the author irrelevant to the build. The highest form of organizational thinking is not the meeting. It is this model that makes the meeting unnecessary.
Forty-two at scale
This was always a risk. AI turned it into an emergency.
In the old world, bad strategy was partly protected by the cost of execution. It took time to build nonsense. The friction of building created a delay between vague thought and wasted effort, and some vague thoughts died or were clarified in that gap before they got expensive.
AI removed the friction. Vague thought now becomes prototypes, backlogs, agents, and dashboards almost instantly. The enumerative document just got a rocket engine. It can produce more artifacts than ever without producing more understanding. The organization starts to confuse generated surface area with progress.
AI also produces new vocabulary at extraordinary speed. First everything had to be promptified. Then MCP-ified. Now agentified. Soon there will be another word. The enumerative document absorbs each one effortlessly, because it was never anchored to a model in the first place. If the strategy is “agentify the workflow,” almost anything can be agentified. If it is “build a governance harness,” almost anything can be called governance. The words move faster than the work. The organization feels current without becoming more capable.
This is forty-two at scale. A machine that returns precise, fluent, beautifully formatted answers to questions no one took the time to define.
The discipline AI demands is the opposite of the one most organizations reach for. When execution becomes cheaper, discernment becomes more valuable. When anyone can generate a plan, a diagram, a demo, the scarce thing is not motion. It is a causal model that points motion toward value. AI should be forcing sharper questions, because the cost of building the wrong thing has fallen. Instead it is being used to move faster through the enumerative — more workshops, more diagrams, more named initiatives, more orchestrated ambiguity. Deep Thought, running at full speed, returning forty-two after forty-two after forty-two.
The builders question
A builder feels all of this immediately. The builder does not want another interpretation layer. The builder wants the problem stated clearly enough that the work can begin, the result can be tested, and the model can be improved. Builders are not anti-change. They are anti-unfalsifiable change. They do not object to revising the model. They object to revising it before it has touched reality.
Causal thinking is not rigidity. It is discipline. It says: think hard enough to make a claim, build hard enough to test it, be honest enough to update it. Enumerative thinking says something quieter and more comfortable: stay broad enough that no one can ever be wrong or challenged.
This is not only a theory of organizations. A team runs on a model or a list. So does a project, a company. At every scale the question is the same. Is the work derived from a view of the world that could be wrong — or arranged from a list of things someone decided to put on it?
When building becomes easy, the old excuse disappears. The hard part is no longer producing artifacts. It is knowing which artifacts deserve to exist. That has always been the builder’s question. Not what can we call it. Not how can we organize it. Not who gets to authorize it. But what is true enough to build on.
Question until you get there.
Then build like maniacs.
And let reality answer back.