The windows to your identity

The two most unique things about you are the patterns of your irises.

Each is formed in the womb in a random process of development that folds, stretches, and weaves together the delicate material around the pupil of your eye.

The result, as computer scientist John Daugman describes it, is, “a meshwork of connective tissue that displays arching ligaments, crypts, contraction furrows, a corona and pupillary frill, coloration, and sometimes freckles.”

Daugman’s research quantified just how unique these are: he calculated the odds of two irises sharing the same pattern are as remote as two people each flipping a coin 249 times and getting identical sequences.

It’s difficult to overstate how unlikely that is.

The probability of a match is just 1 chance out of 2 (heads or tails) raised to the 249th power (the number of flips).

In other words, it’s one chance in a number so unfathomably large it has a seemingly made-up name: a “quattuorvigintillion.”

A quattuorvigintillion is a vigintillion trillions. Or one thousand trillion trillion trillion trillion trillion trillions. Or a 1 followed by 75 zeros.

It’s also just a little less than the number of atoms in the observable universe.

The chance one of your iris patterns will match someone else’s is astronomically small.

Or even that they match each other.

Iris patterns have no basis in genetics, so the irises of identical twins are no more alike than anyone else’s. And your left iris is a complete stranger to your right one.

This makes them the best possible way to definitively tell one person from another.

“Biometric identification systems all rely upon forms of random variation among persons,” Daugman explains. “The more complex the randomness the better, because more dimensions of independent variation produce signatures having greater uniqueness.” 

Irises are complexly random and the most unique things about you.

(I guess I’ll stop pedantically telling people “most unique” isn’t possible now.)

But this is only part of what makes them the ideal identification system.

Biometric identifiers have to be maximally variable between people, of course. But they also have to be minimally variable for the identified person.

For example, facial recognition fails as a biometric identifier because faces can change: the expression you’re making, the angle it’s turned at, cosmetic surgery — even the light you're standing in can make you unrecognizable to a machine.

Daugman notes that, with facial images taken just one year apart, even the best recognition algorithms have error rates as high as 50%.

Irises, by contrast, never change — largely because they’re internal. Sealed behind the cornea, the iris is spared any damage or effect from contact with the outside world. (No cosmetic surgeon would touch them.).

But we can see them, too.

Assuming you’re not lying on an operating table or recovering from a knife fight or something, the iris is the only internal organ that’s externally visible.

That’s only part of what makes them ideal for biometric identification.

Irises are also two dimensional, which means they’re unaffected by either the lighting they’re viewed in or the angle their view at. They’re easy for machines to scan because they have a natural center point (the pupil). And unlike fingerprints, the machine does not need to make physical contact with them.

As a result, Daugman determined that “images of the iris adequate for personal identification with very high confidence can be acquired from distances of up to about 3 feet.”

Like when you’re on a Zoom call, for example, which would be useful.

In 2024, an employee at a British engineering firm wired $25.6 million to a scammer’s account based on a video conference in which every other participant was a deepfake. If Zoom had scanned everyone’s eyeballs, the one human on the call would have known to get off it.

This is not yet possible, unfortunately, but only because the average laptop camera is not capable of capturing the intricate patterns of a human iris. 

But we might soon have the next best thing. Yesterday, Tools for Humanity — the development company behind the World protocol — announced “World ID for Zoom,” an integration that checks a participant’s face on a Zoom call against a photo they took when looking into an Orb — the hardware Tools for Humanity built to confirm you’re a unique human by scanning your iris.

The scan proves you’re a unique human and the selfie proves you did the scan. 

It’s not as fool-proof as scanning the iris directly. But until Apple starts building the Orb into its Macbooks, it’s the best we can do, and probably more than sufficient.

People will have to get less squirmy about getting their irises scanned, though. Only about 18 million people have looked into the Orb so far, in part because putting your eyeballs on blockchain feels a little dystopian.

But what’s the alternative?

As the internet gets increasingly overrun by bots, proving you’re not one will become increasingly important.

Only your irises can do that.

— Byron Gilliam

_{Brought to you by:}

Tokenized money market funds extend familiar liquidity investments into a programmable format, opening new possibilities for automation and streamlined processing.

As markets continue the shift from centralized records to shared ledgers, tokenized money market funds are emerging as a bridge between traditional liquidity and digital assets.

All of finance is coming onchain and the institutional heavyweights are headed to DAS London this November 10-11.

Grab your tickets for unmatched access to the leaders bringing institutional finance onchain.