In this day and age, we have a lot of secrets.

And we're constantly having to give them away to the internet.

Like, Amazon has my credit card number.

I typed it in voluntarily.

So I could order a BSG pendant and some ferrofluid.

But I don't even know where my information is kept.

How can I trust that Amazon doesn't accidentally give it away?

This challenge of secret keeping is an important problem for companies and governments.

That's why encryption, or translating information into a code only the right people can read, is so important-- so important that it was put on the United States Munitions List, along with flamethrowers and bombs, as a weapon regulated for national security.

That was until a student took the United States government to court and encryption was ruled free speech.

Now most of the focus is on improving encryption, because as computers get smarter and faster, these codes become easier to unscramble.

That's why we need to turn to cutting-edge physics to improve encryption.

But let's find out how.

Before I go physics on you, let's talk about codes.

A super simple type of code but one that was used by the Spartan army and probably Scooby-Doo is simple substitution, where you just swap out each letter in the alphabet for a different letter, like so.

For example, DIANNAROCKS would become EJBOOBSPDLT.

This code is easy to generate and almost as easy to crack.

You could even use extra information, like guessing that the six-letter word with the double O's was my name, helping you to crack the code faster.

So cryptographers turned to math to develop harder codes.

They thought a good code should be easy to create but difficult to decode, like a mess.

In math terms, they sought processes called one-way functions-- easy to compute one way but hard in reverse, like multiplying prime numbers.

If I told you to multiply 79 by 73, a few punches on your calculator would give you 5767.

But if I told you to factor 5767, well, you'd have to start by saying, does that I divide by 2 and then 3 and then 5 up through the prime numbers until you finally found 79 times 73.

That would take some key punching.

The largest prime number known today is 2 to the 74,207,281 minus 1, a number that I couldn't even write on the screen if I wanted, because that's more digits than pixels in my HD video.

Now, imagine multiplying that by the second largest known prime number and giving that to your little cousin to factor on a rainy day.

That's just cruel.

But it's great for encryption.

Imagine I want to send the message DIANNA ROCKS to another secret agent, Sophia.

I'm going to use this math.

First, I substitute the letters in my message for bits, like we did in our easy Scooby-Doo code.

If we were using binary, DIANNA ROCKS would look like this.

But we'll just use 1, 2, 3, 4, 4, 3, 5, 6, 7, 8, 9, 10.

The 5 is for the space, and the conversion key from numbers back to letters is publicly known.

No surprises yet.

But then, I multiply each of the numbers by two prime numbers-- say, 11 and 13-- for a total multiplication factor of 143.

DIANNA ROCKS becomes 143, 286, 429, and so forth.

I've told Sofia ahead of time that the key is to divide by 143 and she can decode the message.

Now, say someone intercepts the encrypted message.

They have to figure out what to divide by to decode the message.

This example wouldn't be so difficult.

But what if we used prime numbers that are thousands of digits long?

An eavesdropper might be crunching numbers for longer than they are alive.

This is the idea behind using prime numbers for encryption, though real-life techniques like the RSA algorithm are a bit more complicated.

Now, the limitation with using prime numbers is that we keep building smarter and faster computers that factor out ever more quickly.

Experts warn that with new technologies like quantum computing, codes that currently take hundreds of years to crack could be solved within minutes.

And this is where the physics comes in.

Enter quantum cryptography, a technology that hides information in photons or the particles of light.

Here's how it works.

Say you want to enlighten-- heh, heh-- a chosen stranger on our mantra, or this.

Instead of deciding upfront what the secret multiplication factor or key is, you use quantum mechanics to make one randomly and send it to your recipients.

Here's how the random key is made.

Alice, the message sender, sends photons that are polarized or vibrating in four different directions-- horizontal, vertical, diagonal, left, or right.

Bob, the recipient, measures which direction they're polarized.

Note by using two differently polarized detectors for each photon one at a time back and forth guessing which detector to use randomly, the detectors translate the photons into bits.

Like, a horizontal measurement could register as a 1 and a vertical as a 0 on this detector.

And eventually, Bob will get the multiplication key from this set of bits.

So now, Bob has a measurement of 1, 0, 0, 1, et cetera, for each photon, keeping in mind that he measured them on two different detectors randomly.

Now he compares with Alice.

And for each photon, he'll tell her which detector he used.

And she'll tell him, wrong, right, wrong, right, based on which filter she used to send the photons.

Because, see, Alice sent either vertical horizontal photons or diagonal photons.

And if Bob uses a diagonal detector on a horizontal or vertical photon, according to quantum mechanics, he will have a 50% chance of measuring a 0 and 50% chance of a 1.

That's why Bob's detectors need to match Alice's filters.

After they go through this public check of the order the detectors were used, they throw out each result from when Bob guessed incorrectly.

And now they have a sequence of identically polarized and measured photons.

That sequence is the key.

Alice can now send the actual encrypted message through a traditional channel and use the quantum key to decrypt it.

And mathematicians have proven that if you make a truly random numerical key, you can theoretically make a code called a one-time pad that is unbreakable.

So why can the order of polarized detectors used by Bob be public?

Well, it's not the ones and zeros that he obtains that are being shared.

It's just the order of detectors.

You would still need to send the polarized photons in through those detectors in the correct order to figure out the key.

But those photons were polarized randomly, so the eavesdropper is outta luck.

And things on this scale-- 1,000th of the width of a human hair-- get weird, as they say, because of quantum mechanics.

If an eavesdropper hacks into the system and tries to copy some of the photons using the wrong order of detectors, they'll change the key.

Bob and Alice will know, because they can check for errors in a subset of the bits in the key, and they can try again.

Of course, getting quantum cryptography to work in the real world is not so easy.

Small disturbances can change photon polarization.

And when creating photons, if you get them off by even just a degree, those errors will add up.

Physicists have only been able to send quantum keys over 200 kilometers.

You can even sabotage quantum detectors by shining a bright light on them.

And even if quantum encryption does become commercially viable, much of the internet's infrastructure would have to be rebuilt.

But still, think of how powerful this technology would be.

An eavesdropper has to measure in order to get the key.

But when measured, the key changes.

You can know if you've been hacked even before you send the message because of this fundamental aspect of nature.

The universe is based on probabilities.

Quantum cryptography hides information not by besting a computer but by stowing it within the unknowability of nature itself.

Thank you so much for watching.