What termites and cells have in common | Hacker News Reader

What termites and cells have in common

mpg.de

41 pointsbenbreen13 comments

> One characteristic of self-organizing processes is random fluctuations that can be amplified by local interactions between agents.

It is going to be amazing when we are able to design self organizing organisms and create emergent behavior at will. Our current infrastructure is efficient but fragile, the same applies to software. Even that we are already moving in that direction. Dozens of pods on Kubernetes are more resilient that a more traditional active/passive architecture. Smaller more numerous components controlled by environment status instead of direct control is getting closer to this type of design.

How is the pheromone gradient maintained in the termite hill?

> Seemingly headless

Heedless? Termites have quite well-defined heads.

I interpreted this as without a "head builder"/foreman/architect, which is what makes the structure-out-of-nothing to be a surprise.

There's no head termite in this context, so how are they coordinating such large-scale structure?

Possibly so (I agree there is no "head termite"). But if you observed humans on a building site, I think it might be difficult to identify a "head". They would all appear to be doing their own thing - more or less.

Depends on the level of observation. This is pretty detailed, we’re looking at pheromones and what not. I’d think the equivalent for humans would be looking closely at the communication between them, through the day. I think that would make it obvious there’s a central coordination.

It would be rare to see, for instance, a group of humans spontaneously arrive and assemble a structure without having some pre-coordinated objective. If it's the same structure they've built a hundred times before (say an Amish barn raising), you may not even see any clear or explicit coordination during the work. And on any given day (say when it's being painted) a conventional construction project may similarly appear uncoordinated and headless.

But you can be pretty confident that that morning or some days before what was to be painted and what color was already set out. And when the foundation was laid, and plumbing installed, there was a plan communicated to those workers. Not necessarily the day of, but again sometime in advance.

If we scale to cities and towns, you'll be more likely to see a kind of uncoordinated development. The corner shop wasn't intended to be there on day one (outside of planned cities, at least), but someone recognized that it had value (on a popular foot path, near some homes or offices) and met the need without explicitly coordinating with the others. Of course, this is also tempered by zoning and permitting laws.

To me all of Earth's cities from space look alike, and also like slime mold colonies.

They aren't coordinating. The large-scale structure is an emergent property of their collective behaviour. This phenomenon is most famously illustrated by Conway's Game of Life [1]. Non-local patterns emerge from entirely local rules.

[1] https://en.wikipedia.org/wiki/Cellular_automaton

Does that mean we could also achieve (or actually are) organisation without coordination?

Yes, this is the utopian goal of political anarchism and arguably much market theory. Unfortunately, it's only as good as the right selection of local rules, and it's very possible that what's good for maximizing one thing has many untoward side effects or requires such a degree of homogeneity as to become unpleasantly restrictive.

I know they aren't coordinating, I'm presenting why the author writes about it appearing to be "headless".

I just don't see how a cell is similar to an insect.

Reply on news.ycombinator.com