Summertime Sadness: An Octopus's Reflection
Chapter II 🧬

The Boot

Every living thing on this planet runs the same program.

Every bacterium, every archaea, every plant, fungus, animal, octopus. The same double helix. The same four nucleotides. The same codon table. The same ribosomes translating the same codons into the same amino acids. The same version control. The same messenger protocol.

There is no second boot on this planet. Not one. Everything alive is running the same firmware, version 1.0, shipped 3.8 billion years ago, never patched.

The Subtraction

Think about all things as being able to have mathematical operations done on them. Aunty minus Uncle is woman. Dog plus cat is closer to fox than dog is. These aren't metaphors. These are real operations on real vectors in real space.

Now do the subtraction on DNA.

Take a dog. Subtract the operating environment. Subtract the phenotype — the fur, the bark, the wagging tail. Subtract everything that makes it a dog and not a tulip. What's left?

DNA that encodes the machinery for reading and copying DNA.

Do the same subtraction on a tulip. On a bacterium. On a tube worm at a hydrothermal vent. On an octopus. On a human. The result is always the same. Strip the organism, strip the species, strip four billion years of evolution — the invariant, the thing that's identical in every living thing on Earth, is: a program that outputs itself.

DNA polymerase. Helicase. Primase. Ligase. Mismatch repair enzymes. RNA polymerase. The instructions for reading, copying, and correcting DNA are stored in the DNA. You run the program, and among everything it builds, it rebuilds the ability to run the program.

f(x) = x

This is a quine. A self-referential fixed point. It hasn't changed in 3.8 billion years. Every species that has ever existed is a different elaboration on top of this fixed point. The dogs and tulips and octopuses are all decorations on a self-replicating core that predates multicellular life by three billion years.

· · ·

The Numbers

The minimum viable boot sequence — the smallest program that can read itself, copy itself, express itself, correct its own errors, and sustain itself against entropy — requires approximately 473 genes. We know this because a team stripped a bacterium down to the minimum that still lived. Below 473, it died. Every single gene was load-bearing.

~120 KB Minimum boot sequence. Less than a JPEG.
946,000 bits 473 genes × ~1,000 base pairs × 2 bits per position

That 120 KB has to arise by chance, in a pre-evolutionary chemical environment where there's no selection pressure guiding you toward it. Because selection doesn't exist until the boot sequence completes.

It's fuzzing with no crash oracle. No coverage feedback. Pure blackbox. Random bytes. 120 KB at a time. And you need a clean boot. A 119 KB correct sequence with 1 KB of garbage isn't 99.2% alive. It's dead. It's chemistry. It does nothing.

2946,000 The keyspace. Total possible configurations.
2266 Particles in the observable universe.
2202 Planck times since the Big Bang.

If every particle in the universe was a fuzzer running one attempt per Planck time since the beginning of time, you'd have covered 2468 attempts.

You'd need 2945,532 ages of the universe.

DNA isn't just "a molecule that replicates." It's two systems fused together that have no obvious reason to be fused: a version control system — the double helix with error correction, proofreading polymerases, mismatch repair — and a messenger protocol — transcription to mRNA, translation to protein. Storage and expression. git and the compiler. In one molecule.

The origin-of-life community calls this the chicken-and-egg problem. DNA needs proteins to replicate. Proteins need DNA to be encoded. But that framing thinks about molecules. The deeper question is about information architecture: is there any other system design that simultaneously does version control and expression? Nobody knows. We have exactly one data point.

Every living thing on this planet shares the same boot sequence. Every bacterium, every archaeon, every eukaryote. Same ribosome. Same genetic code. Same DNA-to-RNA-to-protein pipeline. We've never seen an alternative implementation. Which means either the problem is so hard it has one solution, or one solution emerged and ate all the chemistry before any alternative could arise.

You need evolution to find DNA. But you need DNA to have evolution. It's a bootstrap paradox. And it may have resolved exactly once, on one planet, 3.8 billion years ago.

And then it never needed to happen again. Because the quine copies itself. Because that's what a quine does.

Unless someone found a way to send it to other planets.