The Power of Swarms Can Help Us Predict the Future

[Narrator] In 1987, a computer animator

named Craig Reynolds

grew tired of rendering hordes of animals

one by one.

He set out to find a way to automate the process.

He came up with Boids, simple software

to create surprisingly complex group behaviors.

Each virtual individual,

in this case, each paper plane,

is guided by just a few straightforward rules.

With the right set of guidelines

the individuals combine to form a unified flock

moving and reacting together.

Reynolds' flocks were remarkably realistic

making some scientists wonder if real-world swarms

were organized from the bottom up.

Biologist Ian Couzin started looking for rules

that might explain how locusts congregate

into swarms.

Using computer vision software

to map their positions and orientations

he discovered that locusts eat each other

from behind, propelling their neighbors forward

as they try to escape.

Just a single rule, the rule of self-preservation,

ultimately leads to the plagues of locusts

we see in nature.

Now, these rules are being discovered

in all sorts of biological systems.

At his lab at Princeton,

Couzin is applying computer vision techniques

to understand the rules

that govern schools of tiny fish

called golden shiners.

It may turn out that crowds of humans

have more in common with schools of fish

than we ever imagined.