A cubic-convergent iteration for π whose outer recurrence is not hypergeometric — it falls outside RCC's classified class, yet still attains optimal RC₁ cost. Develops derivative engineering and the tower of continua (ℚ → algebraic → analytic → elliptic/modular → periods). Where RCC defines the cost classes, PI_RCC shows their boundary is permeable.

The Simplest Increment

A cubic-speed engine for π, and what it reveals about the architecture of numbers

There is an iteration so simple it looks like a typo:

x → x + sin(x)

Feed it a number reasonably close to π, apply it a few times, and the error collapses at a rate proportional to its own cube. Three correct digits become nine; nine become twenty-seven. The convergence is violent once it begins — and the surprising part is not that it works, but what kind of machine it is, and what that machine tells us about the hidden structure of the real numbers.

This project is a written exploration of that iteration together with a companion verification harness that checks every quantitative and symbolic claim the essay makes. This document is the reader's orientation to both.

What you are looking at

The material comes in two layers, and it helps to know which is which before you dive in.

The essay (the main narrative) is a piece of mathematical exposition. It develops the x + sin(x) iteration from first principles, situates it in a "tower" of number systems, and argues — carefully, and with its own honest caveats — that it represents a small but genuine new kind of computational object.
The companion script is the essay's conscience. It reproduces the symbolic cancellations from scratch in a computer algebra system, iterates the map in exact rational arithmetic, and measures the convergence rates, certificate sizes, and iteration counts the essay predicts. Where the prose says "cubic," the script fits a slope and reports whether it actually came out near 3.

You do not need to run anything to get the ideas; the essay stands on its own. But if you are the sort of reader who trusts a claim more once a machine has checked it, the harness is there, and its final report is organized as a plain table of claim, prediction, measured value, verified? — no ceremony, just verdicts.

The idea, in one breath

Write the error at step n as e = x − π. One application of the map gives, after using the identity sin(π + e) = −sin(e) and the Taylor expansion of sine:

e_next = e + sin(π + e) = e − (e − e³/6 + …) = e³/6 + …

The linear term cancels exactly; the quadratic term is absent by symmetry; what survives is a cubic. That single cancellation is the whole story. It is not a numerical trick — it is a consequence of the symmetry of the sine function about π, the same symmetry that makes every even derivative of sine vanish there.

The essay's central claim is that this cubic behavior follows from three independent conditions on the update function g (here g = sin):

g(π) = 0 — π is a fixed point,
g′(π) = −1 — the linear error term cancels,
g″(π) = 0 — the quadratic error term cancels.

Miss the third and you get a merely quadratic engine. Sine at π satisfies all three at once, and the companion script includes an explicit counter-example — a perturbed sine that keeps the first two conditions but breaks the third — to show the convergence quietly drop from cubic to quadratic. That demonstration is, to me, the most persuasive single experiment in the whole project.

Why it is interesting

A few reasons, in ascending order of ambition.

It is minimal. Cubic convergence usually costs you something — a derivative evaluation, a division, an elliptic-integral step. This map needs none of them. It is arguably the simplest closed-form iteration that converges to π faster than quadratically.
It is a "design principle," not an accident. The essay reframes the whole thing as derivative engineering: you don't stumble onto cubic convergence, you sculpt it by choosing an update function whose value, slope, and curvature line up with the target in a prescribed way. The convergence order becomes a theorem about Taylor coefficients rather than a lucky observation.
It sits at a boundary. The most speculative and (I think) most rewarding part of the essay places this iteration in a tower of number systems — the rationals built from the integers by inversion, the algebraic numbers built by root-extraction, the analytic numbers where π lives — and reads the map as the constructive witness for inverting the sine function at its zero. π, on this reading, is precisely the object you are forced to adjoin when you close the rationals under sine and then ask to run that operation backwards. That "tower of number systems" is the explicit subject of the companion essay The Extension Ladder, which develops the same climb — integers → rationals → algebraic numbers → transcendentals like π → complex numbers — as a general mechanism rather than a single example. Reading the two together, this iteration becomes the constructive witness for exactly the rung The Extension Ladder describes in the abstract: the step from algebraic numbers to the analytic constants that require limits and series to reach.

It is worth being clear about what the project does not claim. It is not a practical way to compute π — Chudnovsky's series remains the champion by a polynomial factor, and the essay says so plainly. The value here is theoretical: a clean, precise example of how mathematical machinery grows by the smallest mutation that existing classifications can't absorb.

A note on intellectual honesty

One thing I want to flag, because it shaped how this material is organized: the project argues against itself in places. An accompanying Socratic dialogue presses hard on the question of whether this iteration truly "certifies" anything or merely "cohabits" with an interesting number, and it lands on a genuinely open verdict. The essay incorporates that pressure rather than hiding from it. If you enjoy watching a claim get stress-tested until only the defensible core remains, you will find that thread running throughout.

Who might find this useful

The mathematically curious reader who likes a short, self-contained result with surprising depth — the cubic-cancellation argument fits on a napkin and still rewards a second reading.
Students of numerical analysis and fixed-point iteration, for whom the three-conditions framing and the quadratic-vs-cubic counter-example are a compact, memorable lesson in why convergence order is engineered.
People interested in the philosophy and structure of the number line — the "tower of continua" and the reading of π as an inverse-of-sine will appeal to anyone who thinks about where numbers come from.
Anyone who wants claims checked. If you distrust prose that asserts "cubic" without evidence, the companion harness exists precisely for you; its report is designed to be read top-to-bottom as a verdict sheet.

Where to go next

Start with the essay and read it straight through; the payoff builds. Then, if you're curious how much of it survives contact with an exact-arithmetic machine, turn to the companion report and read the verdict table. The four headline verdicts to look for are: cubic cancellation confirmed, the fixed point is superattracting (multiplier exactly 0, not repelling), the composite certificate cost lands in the optimal logarithmic regime, and the outer recurrence is genuinely non-classical.

It is not the last word on the structure of the continuum — several floors of that tower are still under construction, and the essay is candid about it. But it is a clean, precise, and surprisingly deep example of how mathematical mechanisms grow, one increment at a time. Enjoy.

PI_RCC — The Simplest Increment: x + sin(x)