We've all seen people act in ways that seem totally irrational —and for most of us, unpredictability is just part of human nature.

But for scientists, when something is unpredictable, that's usually a sign that they don't fully understand what's going on.

Which isn't really shocking when it comes to people.

Our brains are just vastly more complex than we have the tools to understand right now.

But even if we can't understand the brain down to the deepest level,

some psychologists think that a set of ideas borrowed from quantum physics could help us make sense of human behavior.

The notion is called quantum cognition—and it isn't suggesting that our brains actually function at the quantum level,

just that the mathematical tools of quantum mechanics could help make human behavior more predictable.

Now, the idea that this bizarre branch of physics could be useful for understanding the brain might seem like a stretch—

but the reason it's useful is because quantum mechanics is all about statistics.

For instance, it's impossible to know where an electron is at any given time;

you can only know how likely it is that, when you measure it, you will find it in a given place.

And that's not because we're bad at measuring—that's just how the universe works on a fundamental level.

Statistics are also useful in other branches of science, though, for different reasons.

They can help us understand the big picture even when we don't know all the lower-level details.

Like, you can use statistics to predict how a group of people will vote even if you don't know what every individual person will do.

And when it comes to the brain, there are a bunch of low-level details that we don't understand.

For example, while neuroscientists have figured out where our short-term and long-term memories are stored,

it's still not clear how your brain selects certain details to remember and others to forget.

But in some cases, we can use statistics to make decent predictions about how people will behave even if we're not really sure why.

Traditionally, these cognitive models have relied on classical probability, which can be pretty black or white.

Like, you either have three aces in your poker hand or you don't.

You either blew off that important project or you didn't.

Or did you just kind of drag your feet until it was too late?

See, human cognition is full of ambiguities, and classical probability just isn't well-suited to handling them.

That's where quantum mechanics might be useful.

The quantum world is anything but black and white, but we've developed really powerful tools to deal with that ambiguity.

And recently, psychologists have been exploring whether or not those tools could be reused to help understand the brain.

In fact, quantum cognition models are already performing as well or better than classical models at predicting some kinds of human behavior.

For example, it's really tricky to predict how humans will make decisions.

Like, no matter how well you know someone, the things they decide just don't always seem logical.

In the early 1990s, psychologists showed this with a simple experiment.

In it, they asked 98 subjects to guess the result of a coin flip.

If they were right, they'd win 200 dollars, and if they were wrong, they would owe 100.

After the first flip, everyone was asked if they wanted to play again.

In general, both winners and losers wanted to flip again,

which is reasonable since you're more likely to win money after multiple rounds than lose it.

But that was not true for everyone.

Specifically, subjects who weren't told whether they'd won or lost the first coin flip mostly decided not to play again.

Even though the odds were in their favor.

From the perspective of classical cognition models, this doesn't make any sense.

After all, if subjects were told whether they'd won or lost, neither result affected their decision—in general, they all wanted to play again.

This is called the sure thing principle because all the options seem to lead to the same result.

But somehow, not knowing made it not a sure thing.

And while classical models of cognition struggle to explain how that could be, one principle from quantum physics does offer a way to understand it.

See, in the quantum world, just the fact that something is unknown can change the outcome of an event.

That's the premise of the famous double-slit experiment.

In one version of the experiment, physicists fire a beam of electrons at a detector.

In front of the detector there's a barrier with two slits in it.

When the beam is turned on, the electrons strike the detector in a pattern

that looks a lot like the pattern you get when two sets of ripples interfere with each other.

The weird thing is, even if you release the electrons one by one, you still get this interference pattern.

In other words, the electron is interfering with itself.

That's because it doesn't exist in a single, precise location, so there's some ambiguity about which slit it passes through.

But, if you set up a sensor to measure which gap each electron travels through, this diffraction pattern disappears.

So, the instant the electron's position is known, the ambiguity is gone, and the interference is, too.

So, broadly speaking, this shows that in quantum mechanics, simply not knowing can produce a totally unexpected result.

Similarly, in the coin-flip experiment, just the existence of doubt changed the likelihood that someone would play another round.

As illogical as these scenarios sound, though, neither one is unpredictable.

In physics, scientists use quantum probability theory, which is a model that accounts for the fact that knowledge of something can affect the result.

And weirdly enough, you can apply the same theory to human decision-making to predict how people will make decisions,

even if we don't understand precisely why.

Psychologists were able to use this quantum model to correctly predict people's decisions in the coin-flip experiment, even when the classical model failed.