Thread by @SarahJamieLewis

Seems a few communities are explicitly basing their consensus algorithms off of various insect models. Who wants a cryptocurrency thread that is actually a thread about ants?

Let me tell you about alarm pheromones.

The impact of pheromones on the behavior of ants (and other insects) has been well studied for about 50 years now. Mostly by researchers poking ants with sticks and the poking other ants with the same stick, or poking the stick into some goo and then poking the ant with the stick

We actually know a heck of a lot about the various chemical receptors that ant species as developed, but there is one that I find very interesting and that is the branch of chemicals that act as "alarm pheromones"

The evolutionary nature of alarm pheromones is pretty easy to derive, ants in danger excrete a chemical, this chemical either calls other ants to help or makes ants run away from danger.

Both very useful evolutionary traits to develop.

One less than predictable evolutionary result of the development of alarm pheromones is the species commonly referred to as "slave-making ants"

These ants are characterized by their reliance on stealing broods from other ant species to survive.

"Of the approximately 10,000 species of ants, only a tiny minority of about 50 are active slave makers. However, slave making apparently has evolved independently more than ten times" (dettorre 2001)

Many slave-making ant species have adapted the alarm pheromones from a defensive purpose to an attack purpose - commonly referred to as "propaganda pheromones"

They excrete these alarm chemicals in much larger quantities, triggering an overload in the colonies they attack.

This causes the ants in the nests they attack to become confused, many will start attacking their own nestmates.

During this confusion the slave-making ants steal larvae and bring in back to their own nets.

The larvae then grows up and perform functions that the slave-making ants are either unable or very inefficient at performing (colony maintenance, brood care, foraging).

That's what decisions based on local signalling get you.

This is certainly not the only documented case of pheromones being subverted by evolution. Similar behavior has been observed in bees, where social parasite species use chemical deception to steal food despite a whole host of colony defense mechanisms.

And that's the thing about local signalling, it's really cool, and remarkable, but we have hundreds of examples of species that have evolved to exploit it.

Exploitation can be entirely destructive (as in slave-making ants) or piecemeal theft (as in stingless bees)

(Not to mention the fact that most of these local signalling processes only work because the environment and evolution have forced density and population parameters into certain bounds - when we attempt to bring these into a digital domain without those bounds, they fall apart)

Anyway, this is actually a thread about why your insect inspired consensus mechanisms won't magically build robust decentralized consensus.

Also, since this is a thread on ants, It would be terrible if I didn't point out that ants are basically one big fuck you to everything you think you know about how nature works.