What Is the Riemann Hypothesis and Why Is It Unsolved?

Why don't the prime numbers follow a pattern?

2, 3, 5, 7, 11, 13, 17, 19, 23... Look at the gaps. 1, 2, 2, 4, 2, 4, 2, 4, 6. No rule tells you where the next one lands. You just have to check.

And yet, zoom out far enough, and the primes stop looking random at all. Their overall count obeys a beautifully smooth law. Something is holding this chaos together from a distance.

The Riemann Hypothesis is a guess about exactly what that something is. It's been open since 1859. There's a million-dollar prize for solving it. And it all comes down to one question: does every "zero" of one specific function sit on one exact line?

Let's build up to what that means.

Primes look random up close. Do they, though?

Here's a claim: even though you can't predict the next prime, you can predict how many primes show up below any number $N$, at least approximately.

The count of primes up to $N$ is written $\pi(N)$. So $\pi(10) = 4$, because 2, 3, 5, 7 are all the primes up to 10.

Carl Friedrich Gauss noticed, as a teenager, that $\pi(N)$ tracks the curve $N / \ln(N)$ remarkably closely. Not exactly. But the ratio between them creeps toward 1 as $N$ grows. This became the Prime Number Theorem, proven over a century later.

Drag $N$ up. The jagged staircase (the real count) never matches the smooth curve exactly, but it never wanders far either. A law lives underneath the noise.

But here's the thing: knowing there's a law and knowing exactly how tight it is are two different problems. The Riemann Hypothesis is about the second one. It asks precisely how much the staircase is allowed to wobble away from the curve, forever.

Where does a "smooth law for primes" even come from?

This is the strange part. The tool that pins down the primes isn't a formula about primes at all. It's a function that, at first glance, has nothing to do with them.

Start with a sum. Add up the reciprocals of every whole number, each raised to a power $s$:

This is the Riemann zeta function. Leonhard Euler had already found something remarkable about it decades before Riemann: you can write the exact same sum as a product, using only prime numbers.

Read that again. The sum runs over every integer. The product runs over only the primes. They're equal. The primes are secretly baked into the zeta function's DNA, because every integer factors uniquely into primes, and this formula is just that fact, dressed up in an equation.

Watch the partial product creep toward $\zeta(2) = \pi^2/6$ as you add primes one at a time. It never needs a non-prime. Primes alone rebuild the whole function.

That's the hook. If the zeta function is made of primes, then studying the zeta function should tell us something about how primes are distributed. And it does, through its zeros.

What's a "zero" of this function, and why do we care?

A zero of $\zeta(s)$ is just a value of $s$ where $\zeta(s) = 0$. Plug in a negative even number, like $s = -2$, and you get zero. These are called the trivial zeros. They're a known, boring, understood family. Nobody worries about them.

But the sum defining $\zeta(s)$ only converges, in the ordinary sense, when $s$ has real part greater than 1. So how does it even make sense to plug in $s = -2$?

Mathematicians extend the function using a technique called analytic continuation. Think of it like this: you build an instrument that only works in a narrow range, then discover its readings extend smoothly, in only one consistent way, to a much wider range. The extended zeta function agrees with the original sum wherever the sum works, and it's the unique smooth continuation everywhere else, including at negative numbers and, crucially, inside a vertical band called the critical strip, where the real part of $s$ is between 0 and 1.

Inside that strip live the non-trivial zeros. Nobody has ever found one off a single specific line. Nobody has proven there can't be one.

The Hypothesis itself

Here's the claim, in full:

Every non-trivial zero of $\zeta(s)$ has real part exactly $\frac{1}{2}$.

Picture the complex plane. The critical strip is a vertical band from $\text{Re}(s) = 0$ to $\text{Re}(s) = 1$. Down the middle runs the critical line, $\text{Re}(s) = \frac{1}{2}$. Riemann's guess is that every single non-trivial zero, and there are infinitely many, sits exactly on that one line. Not near it. On it.

To feel why this is such a sharp claim, think of $\zeta(\sigma + it)$ as a radio. $\sigma$ (the real part of $s$) is a tuning knob. $t$ (the imaginary part) is a frequency you sweep through. $|\zeta(\sigma + it)|$ is the signal strength coming out of the speaker.

Set the knob anywhere except $\sigma = 0.5$ and sweep through frequencies. The signal wobbles, dips, gets quiet sometimes... but it never truly goes silent. Set the knob to exactly $\sigma = 0.5$ and hit play. Now the signal drops to true zero, over and over, at very specific frequencies. Those frequencies are the heights of the zeros: 14.13, 21.02, 25.01, and on forever.

Every zero anyone has ever computed, and that's trillions of them, sits on that line. Nobody has found a single exception. Nobody has ruled one out either.

Why should you care where some zeros sit?

Because those zero locations directly control how bumpy the prime staircase from earlier is allowed to be, forever, no matter how far out you go.

If every non-trivial zero sits on the critical line, a mathematician named Lowell Schoenfeld showed exactly how tightly the prime count has to hug its smooth prediction:

where $\text{Li}(N)$ is a slightly more refined smooth estimate than $N/\ln(N)$. Without the Hypothesis, the best proven bound is far weaker, and it gets relatively worse the further out you go.

Drag $N$ out to a million. The real error (orange) stays comfortably inside the bound the Hypothesis promises (blue), just as it has for every value anyone has ever checked. That's the entire stake of the problem: is this snug fit a coincidence of small numbers, or a law that holds forever?

So why is it still unsolved?

Because "true for every number ever checked" and "true for every number, period" are completely different kinds of statement. You can check zeros forever and still never rule out the possibility that zero number $10^{100}$ breaks the pattern.

Riemann proposed the Hypothesis in a single nine-page paper in 1859, almost as an aside while proving something else. David Hilbert listed it among his 23 defining problems for mathematics in 1900. It's one of the seven Millennium Prize Problems, each worth a million dollars from the Clay Mathematics Institute. Trillions of zeros have been checked by computer, every single one on the line. And still, nobody has found a proof that it must always be true, and nobody has found the single counterexample that would prove it false.

Whole branches of number theory currently rest on statements of the form "assuming the Riemann Hypothesis...". If it's ever proven, a huge amount of already-published mathematics locks into place as certain. If it's ever disproven, just as much of it unravels.

The short version

The prime numbers look chaotic one at a time, but their overall count obeys a smooth law. That law is controlled by where the zeros of the Riemann zeta function sit inside a narrow strip of the complex plane. The Riemann Hypothesis says every one of those zeros sits on one exact line, right down the middle. If true, it means the primes are distributed as evenly as they possibly could be, with an error that never gets out of hand. Every zero ever found agrees with the Hypothesis. Nobody can yet prove it holds for all of them, and that gap between overwhelming evidence and actual proof is why it's stayed unsolved for over a hundred and sixty years.

All visualizations are interactive React components computed live in your browser, including a direct numerical approximation of the Riemann zeta function via Euler-Maclaurin summation. No libraries beyond React.