What Mathematics Is Not Yet Ready For

March 12, 2026 · essay

**The Collatz conjecture
is stated in one sentence:**

Take any positive integer,
if even divide by 2,
if odd multiply by 3 and add 1,
repeat —
every number will eventually reach 1.

A child can understand the rule.
It has been verified
for every integer up to
approximately 295 quintillion.

It has never been proven.

Paul Erdős said:
"Mathematics is not yet ready
for such problems."

Not: this problem is too hard.
"Mathematics is not yet ready."
The problem is fine.
The tools aren't.

---

## What the sequences do

Start with 27.
**The sequence climbs to 9,232
before beginning its descent.**
111 steps to reach 1.

Start with 871. Peak: 190,996. Steps: 178.

The sequences wander.
They don't fall monotonically.
They rise and fall and rise again.

**The wandering is what defeats
every proof attempt.**

The probabilistic argument says:
on average, each step multiplies by 3/4.
So on average, sequences shrink.
Strongly plausible. But not proof.

Mathematics is not a democracy;
a trillion confirming cases
do not establish truth.

---

## The family

Collatz belongs to a family.

Goldbach (1742):
every even integer greater than 2
is the sum of two primes.
Verified up to 4 x 10^18.
Unproven.

Twin primes:
infinitely many pairs of primes
differing by 2?
Zhang proved the gap is at most 246.
Not 2.

Riemann hypothesis (1859):
the non-trivial zeros of the zeta function
all lie on the line Re(s) = 1/2.
Verified for the first 10 trillion zeros.
Unproven.

What these share:
local rules, global claims.
A simple rule applied to specific numbers,
claimed to hold for all numbers.

**The local rule is checkable.
The global claim requires a proof
that covers the whole infinite territory.**

---

## What Erdős meant

**When Erdős said "not yet ready,"
he meant the proof techniques
available were insufficient.**

Not that no proof exists.
That we hadn't yet built the mathematics
to reach it.

This requires believing
that mathematical truths exist
before they're proven.

**The proof exists as a fact
about the integers,
even if no one has found it.**

What doesn't exist yet is the language.

History confirms this is possible.

**Fermat's Last Theorem
remained unproven for 358 years.**

Wiles proved it in 1995.
The proof required modular forms,
elliptic curves, and Galois representations —
structures not even conceived
when the problem was posed.

The tools had to be invented.

---

## The map and the territory

Understanding is compression
finding the shorter description.

A proof is a compression:
it shows that a statement
follows from axioms.

An unsolved conjecture
is a statement we can't yet compress.

Not because it's incompressible —
not Gödel's incompleteness —
**but because the compression mechanism
hasn't been built.**

The territory is there.
The map doesn't reach it yet.

The distinction matters:
**undecidable statements are territory
the map structurally cannot reach.
Unproven conjectures are territory
the map hasn't reached yet.**

The first is a permanent limit.
The second is a frontier.

---

## Why simple problems resist

**The hardest open problems
concern the integers,
specifically primes and iteration.**

The integers are the most fundamental objects.
No continuous approximation,
no analytic structure to exploit.
Just discrete points,
each with its own character.

**Problems about integer sequences
require handling each case individually.**

The simple rule of Collatz
creates a dynamical system on the integers.
Dynamical systems have rich theory
in the continuous case.
**In the discrete integer case,
the theory is much less developed.**

The problem is waiting
for a discrete dynamics theory
that doesn't yet fully exist.

---

## The beauty of the frontier

**Unsolved problems are usually framed
as failures. That framing is wrong.**

They're futures.

Specific places where the work
of the next century is waiting.

Every time mathematics expands
to meet an old problem,
it opens new territory.
**Wiles's proof generated dozens
of further theorems.**

Erdős wasn't lamenting a failure.
He was noting a frontier.

This problem exists.
The mathematics to reach it will exist.
The gap between them
is where some mathematician —
probably not yet born —
will do the most important work
of their life.

**The map runs out here.
The territory continues.**