To the outside world, a life spent in pure mathematics must look like a form of madness — a quiet, intellectual rendering of Camus's Sisyphus, a solitary figure shouldering a massive immovable stone up a vertical cliff, only to watch it slip and roll the whole way back down to the valley.
Look closely at the struggle over the Riemann Hypothesis, though, and the number theorist is not the tragic figure in that painting. They are the absurd hero of it.
This is a companion to The 1,000-Year Nap and the Longshoreman's Desk and Possessing the Proof.
01
The hypnotic rock and the sideways slope
Why the modern number theorist stopped pushing straight up
Riemann conjectured in 1859 that every nontrivial zero of the zeta function lies on a single vertical line in the complex plane, where the real part is exactly one half. For more than a hundred and sixty years the strongest minds available have put their shoulders to that stone and pushed it directly at the summit of a head-on proof.
The mountain has beaten all of them, and it does not beat them by looking hard. It beats them by looking easy. James Maynard has said the problem almost sucked you in, and that it looked far more innocent than it was. That is the specific character of the trap: the statement is short, the object is elementary, and the difficulty is entirely structural.
What the modern mathematician does next is delightfully absurd. Unable to reach the peak, they decline to despair and instead come at the mountain sideways. If the hypothesis will not be proved outright, then prove that the exceptions to it are strictly bounded. Take the foothold on the cliff face and call it what it is — progress.
The 2024 Guth–Maynard result is the cleanest example on record. They did not reach the summit. They translated the zero-density question into Dirichlet polynomials, built matrices, recast the problem as a hunt for a largest eigenvalue, and then played a chess gambit: they deliberately refused the standard simplification, letting the mathematics look bulkier and more chaotic in the short term in order to expose a deeper symmetry that the tidy version had been hiding.
Maynard is candid that these are probably not the techniques that will finish the job — the hypothesis itself will need some big idea from somewhere else. And yet they climbed higher than anyone had in over eighty years. The value was never a static victory. It was the creative, defiant climb.
It almost sucked you in, and it looked much more innocent than I think it was.
- Established — The 2024 Guth–Maynard improvement on zero-density estimates, via large-value bounds for Dirichlet polynomials, broke Ingham's 1940 record.
- Established — The result bounds the possible exceptions to RH; it does not prove RH, and its authors say so explicitly.
- Analogical — Reading the refusal-to-simplify step as a 'chess gambit' — a description of the strategy, not of the technique.
02
The problem that fights back
Where Camus meets Hein: the rock as active opponent
This is the point at which the absurd joy of Sisyphus meets the playful logic of Piet Hein. Sisyphus's rock is not a passive lump of granite. It is an opponent with a position. Hein put the whole relationship in four lines: problems worthy of attack prove their worth by fighting back.
An easy mountain is not worth climbing. The Riemann Hypothesis is the ultimate worthy opponent precisely because it fights back with elegance rather than mess — it defeats each new method on structural grounds, which is why every serious attempt has had to invent new analysis, new logic, new geometry to get any purchase at all. The failed attacks built most of analytic number theory as a by-product.
That is the honest ledger of the last century and a half. The stone has not moved to the top. The people pushing it have been comprehensively rebuilt.
The struggle itself toward the heights is enough to fill a human heart. One must imagine Sisyphus happy.
- Established — Substantial areas of analytic number theory — sieve methods, zero-density estimates, random matrix connections — emerged from attempts on RH and related prime questions.
- Licensed inference — That the by-product exceeds the target in value. Defensible on the historical record, not provable.
- Analogical — Treating the problem as an agent that 'fights back'. Hein's own framing, and a metaphor throughout.
03
The Gödelian loop of the infinite push
Robin's inequality, Π⁰₁ form, and the trapdoor in the logic
There is a logical climax to the endless loop, and it is stranger than the metaphor that got us here. Robin's theorem establishes that the Riemann Hypothesis is equivalent to a simple computable arithmetic inequality holding for every integer above 5040. That makes RH a Π⁰₁ statement: equivalent to the claim that a specific program, checking one integer after another forever, never prints NO.
The consequence is a genuine trapdoor. If RH is false, a counterexample is a finite, computable object. The program would eventually reach it, print NO, and the hypothesis would be refuted in finitely many steps — which means falsity is, in principle, always demonstrable.
Turn that around. If RH can be shown independent of a sound arithmetical theory such as ZFC — neither provable nor refutable in it — then no counterexample exists, because a counterexample would have furnished a refutation. Independence would therefore entail truth. Unprovability, in this one unusual case, is a proof.
Two guardrails, because this is where popular accounts overreach. The argument requires the theory to be sound (arithmetically correct), not merely consistent; and independence is not a thing anyone has established, only a possibility Martin Davis and Paul Cohen both took seriously after long, failed attacks. What is established is the shape of the statement, and the shape is what makes the paradox available at all.
The image survives the caveats intact. Like Sisyphus, the calculation must run to infinity without ever reaching a closed end. Being condemned to push the arithmetical stone forever would be the strongest available testimony to the truth of what we are pushing.
Problems worthy / of attack / prove their worth / by fighting back.
- Established — Robin's criterion: RH holds iff σ(n) < e^γ · n · ln ln n for all n > 5040 — putting RH in Π⁰₁ form.
- Established — For a sound theory, independence of a Π⁰₁ statement implies its truth, since a counterexample would be a finite refutation.
- Asserted — That RH is independent of ZFC. Nobody has shown this; it is a live speculation, not a result, and the 'unprovable therefore true' headline is conditional on it.
- Licensed inference — Reading the mathematician's condition as Camusian rather than tragic — an interpretive frame, offered as one.
04
We must imagine the mathematician happy
The defiant smile at the bottom of the critical strip
From the outside, a life spent on pure mathematics must look like a quiet form of madness — a solitary figure straining at an immovable boulder on a vertical cliff, watching it slip and roll back down to the valley, and then walking down after it.
Look closely at the actual behaviour, though, and the number theorist is not the tragic figure in that picture. They are the absurd hero of it: the one who has fully registered that the summit may be unreachable, declined both despair and false hope, and gone back to the stone anyway because the climb is where the meaning is kept.
They stand at the bottom of the mountain, look up at the towering, beautiful critical strip, adjust their glasses, and with a defiant smile put a shoulder right back to the stone.
Four grooks for the climb
Written in Hein's ironic register: the eternal struggle, the hidden harmony of the Soma Cube, and the playful sovereignty of a mind that declines the committee.
Grook
The Sisyphean Sum
On the beauty of a problem that fights back
We push the arithmetic stone,
One integer by integer,
Up toward a peak that's never known,
Where formulas are simpler.
But when the boulder slips your grip,
Don't stand in sorrow, moping:
It was the struggle of the trip
That kept your spirit hoping.
Grook
The Soma Freak
On the world's smallest philosophical system
Seven jagged, awkward shapes,
Each asymmetrical and wild,
From which a tidy grid escapes
Like chaos from a child.
Yet spin them sideways in your hand,
And let their jaggedness align:
A perfect cube will suddenly stand,
A chaos captured by design.
Grook
The Pests of Proof
On the institutional gatekeepers of truth
Those who lock the temple gate,
And claim they hold the final key,
Will sit in ceremonial state,
And blind themselves to what they see.
But truth is found by those who stray,
On docks, or paths where starlight shines,
Who throw the drilled-in rules away,
To draw their own celestial lines.
Grook
The Recursive Bounce
On playing mental ping-pong with an AI
To play a game of mental chess,
Or ping-pong with a phantom ball,
Is simply learning to confess
That none of us can master all.
You bat a pattern o'er the net,
The phantom spins it back to you—
And in the bounce, you suddenly get
A completely sideways view.
What would show this wrong
- Produce a direct proof or disproof of RH by conventional analytic means, and the 'sideways slope' reading of section 01 becomes a description of a detour rather than a method.
- Show that the Guth–Maynard refusal to simplify was in fact the standard route in disguise, and the gambit framing collapses into ordinary technique.
- Exhibit an arithmetically unsound theory in which the Π⁰₁ trapdoor argument is run without qualification, and section 03's guardrail is doing real work — the popular 'unprovable therefore true' claim fails there.
- Establish RH as provable in Peano Arithmetic (as Broughan argues), and the independence speculation — and with it the whole Gödelian loop — drops out of relevance.
- Demonstrate that the historical by-products of failed RH attacks were developed independently of those attacks, and section 02's central claim about the value of the struggle is overstated.
Sources
- Albert Camus, The Myth of Sisyphus — The absurd hero and the closing imperative quoted in section 02.
- Guth–Maynard and the density hypothesis — The 2024 improvement on Ingham's 1940 bound via Dirichlet polynomial large-value estimates.
- Robin's theorem — The divisor-sum inequality above 5040 that puts RH in Π⁰₁ form.
- Arithmetical hierarchy — Π⁰₁ statements — Why a false Π⁰₁ statement is always refutable in finitely many steps.
- Piet Hein (scientist) — The grooks, the Soma cube, and the 'universal pest' line.
- Soma cube — Seven irregular polycubes, 240 distinct solutions, and Hein's unity-from-variety framing.