Skip to content

The Diamond Lock: Writing Notes a Future Robot Can't Read

Quantum computers will slice through today's internet locks like a laser through glass. Inside the race to build math even a future super-machine can't crack: public-key crypto, Shor's algorithm, and the diamond lock.

planted June 19, 2026 · last tended June 23, 2026

A captured spark. Unverified, unpolished, possibly wrong.


Somewhere right now, a stranger is hoarding secrets they can’t read. Encrypted files they have no key for, scraped off the wire and filed away to age like wine. They’re betting that in ten years they’ll own a machine that pops the cork. The play even has a name, the kind that should cost you a little sleep: Harvest Now, Decrypt Later. Steal the locked box today. Come back for the key from the future.

Here’s the thing. That’s a rational bet.

Every time you log in, every time your mom buys something online, your computer drops the data into a glass box and snaps it shut. The textbook name is public-key cryptography, and its most famous flavor, RSA, hides your secret behind a multiplication problem. Multiplying two enormous primes together is trivial, a rounding error of a second. Running that backwards, cracking the answer into the two primes that made it (the move called factoring), is so brutal that a normal computer would still be grinding away long after the sun burns out. To anyone holding a crowbar, the glass might as well be a foot of steel.

Then quantum computing strolls up with a laser.

A quantum computer doesn’t whack the box. It dissolves the wall the box is made of. An ordinary bit is a 0 or a 1, one or the other, end of discussion. A qubit refuses to commit: it holds a blurry both-at-once that physicists call superposition, which lets the machine wade through a mountain of possibilities in a single parallel breath. In 1994 a mathematician named Peter Shor wrote down the recipe. Shor’s algorithm hands a big-enough quantum computer the exact trick RSA prayed nobody would find: fast factoring. The glass doesn’t shatter. It melts.

So the world’s cryptographers are sprinting to forge a better lock before the lasers finish booting. The field is post-quantum cryptography, and the whole idea is to quit betting on multiplication and start betting on problems a laser can’t carve. The front-runners are lattice problems: picture hunting for one exact point in an infinite grid tilted across hundreds of dimensions, where every direction you turn is the wrong one. In 2024 NIST (the standards body that referees this stuff) crowned the first winners, with sci-fi handles like CRYSTALS-Kyber. That’s the diamond lock. And we’re bolting it onto the internet right now, while the burglars are still unpacking their gear.

This is the attacker’s game played from the defender’s chair (the-attackers-mindset-is-systems-thinking). You don’t win by reacting to the break-in. You win by picturing the burglar ten years out, the one with the laser, and out-thinking a person who doesn’t exist yet. Lean on that idea hard enough and it quietly turns into philosophy (philosophy-the-why-game): is any secret ever truly unknowable, or only unknowable for now?

Which leaves one uncomfortable thing to sit with. If the laser-burglars are already half-built, and the message you sent this morning is parked in someone’s vault waiting its turn, then the only secrets that outlive the decade are the ones you sealed with diamond today. So pick one. That note you fired off an hour ago: glass, or diamond? And how sure are you that nobody’s already got a copy, watching the power light blink on?

Paths that lead here

No paths yet. This note is still off the beaten track.

Where this note points

More from these beds