First-principles thinking

The debate is not the problem

Most writing about hard subjects inherits its terms from the debate already in progress. The question arrives pre-loaded — how do we restore trust in government, how do we make schools more rigorous, how do we keep AI safe — and the writer's job is assumed to be picking a side within a frame someone else built. This is the single most reliable way to produce something that sounds informed and changes nothing, because the frame is usually where the error lives. The debate about why states fail is conducted almost entirely in the vocabulary of character — corruption, laziness, missing courage — and that vocabulary is itself the mistake. First-principles thinking is the discipline of refusing the inherited frame: stripping the problem back to what is physically, economically, and structurally true, and rebuilding the argument forward from there, even when the rebuild lands somewhere the debate has no words for.

This is the move that opens ENSI's systematic analysis of the Czech state's malfunction. The conventional frame says a malfunctioning state is a state with bad people in it. The first-principles move asks a flatter question: what does governing a complex society physically require, and which of those requirements is actually missing? The answer that falls out — that the binding constraint is throughput, the rate at which the system can analyze, decide, and execute, not the virtue of the people doing it — is invisible from inside the character frame and obvious once you rebuild from the function. The article is persuasive not because it argues harder than the moralists but because it refuses their starting point. It went back to what a state is before deciding what a state's failure means.

Strip to the physics, rebuild forward

The method has a specific theory of what "first principles" means in practice, and it is not vague reductionism. It means finding the layer of the problem that is not itself a matter of opinion — the unit economics, the labor content, the latency, the cost curve, the information flow — and treating that layer as bedrock. Once you stand on something true, you rebuild upward, and the rebuild is where the surprising conclusions come from, because you are no longer constrained to land on a conclusion the debate already contains. ENSI's decision complexity work is this discipline applied to the act of deciding: rather than accept the folk theory that good decisions come from good judgment, it decomposes a decision into its actual load-bearing variables — interdependence, stakeholder conflict, data quality, time horizon, stakes — and rebuilds the concept of decision quality as a function of how much of that load the decider can hold. The conclusion that you should match analytic method to measured complexity is not available to anyone still arguing about whether leaders are wise.

The same descent-to-bedrock drives the economic pieces. The economics of infinite intelligence does not ask the inherited question of whether AI will help or hurt business; it strips the situation to a single physical fact — the marginal cost of cognitive work is collapsing toward zero — and rebuilds the entire argument from that one true thing, arriving at conclusions about scarcity, margin, and business models that the surface debate never reaches. Theoretical pillars of economic reasoning is, in effect, an inventory of which economic claims are bedrock and which are inherited convention. And brain vs LLM: the similarities and differences refuses both the breathless "it thinks like us" and the dismissive "it's just autocomplete" by going back to the mechanisms and rebuilding the comparison from what each system actually does. In every case the structure is identical: discard the framing, find the floor, build up.

Where the floor is, and how you know

The hardest part of first-principles work is not the rebuilding; it is locating the floor — knowing which assumption is genuinely bedrock and which is just a convention wearing the costume of a fact. The method's test is whether an assumption could be otherwise: if a "fact" is really a choice the field made and could un-make, it is not bedrock, and the analysis must descend past it. Measuring intelligence performs exactly this audit on a concept everyone treats as settled, showing that much of what passes for a measurement of intelligence is a convention about what to count, not a fact about the world — and that once you strip those conventions away and rebuild, intelligence looks like the integration of complexity rather than a score. Intelligence is complexity integration is the rebuilt structure that audit produces. The discipline is recursive: you keep descending until you hit something that could not be otherwise, and only then do you turn around and build.

This is why first-principles thinking is generative rather than merely critical. It would be easy to read all of this as a debunking habit — a reflex for telling people their frames are wrong. But the destruction is only the entry fee; the value is in the forward rebuild, the new structure that stands on the recovered bedrock and supports conclusions the old frame could not bear. ENSI's pieces do not stop at "the character story is wrong about the state." They rebuild all the way to a buildable alternative — throughput as infrastructure, decision quality as a tunable function, intelligence as something a society can own and run. The first-principles move earns the right to be constructive precisely because it has gone down to the floor and verified that the new structure rests on something true. Anyone can contradict the consensus; the method's standard is higher, which is to replace it with something load-bearing.

Why this has to come first

First-principles thinking is the first of the six because every other part of the method depends on it. You cannot synthesize across disciplines if you are still trapped in one discipline's framing; you cannot reframe contrarian-style without first having stripped the consensus to find its load-bearing assumption; you cannot weigh evidence honestly while the inherited frame is silently deciding which evidence counts. The descent to bedrock is what clears the ground for everything else. It is also what makes the writing feel less like commentary and more like engineering — the sense, which readers of the Czech malfunction analysis or the economics of infinite intelligence report, that the argument was derived rather than opined. That sensation is the trace of a rebuild from first principles.

The wager of this principle is that the most valuable thing a writer can do is not argue well inside the frame but notice the frame and go beneath it. The debate about character was conducted brilliantly for centuries and was wrong the whole time, because no one questioned the floor it stood on. ENSI's method starts by questioning the floor — every time, on every subject — and rebuilding only on what survives. Strip the problem to what is actually true. Rebuild forward. The conclusions that frighten the debate are usually just the structure that was always there underneath it, waiting for someone to stop arguing long enough to look.

Further reading

• Czech State: A Systematic Analysis of Malfunction
• Decision Complexity
• The Economics of Infinite Intelligence
• Theoretical Pillars of Economic Reasoning
• Brain vs LLM: The Similarities and Differences
• Measuring Intelligence
• Intelligence Is Complexity Integration