5 ms·
Because you're curious you get my curious answer. Philosophically I would argue yes, all organisms are capable of swarming. I do not think any species has been
by odomojuli 6y ago
Because you're curious you get my curious answer.
Philosophically I would argue yes, all organisms are capable of swarming. I do not think any species has been able to sufficiently adapt or evolve unless in copious and sufficient populations which requires some kind of implicit understanding of sheer numbers. Swarming only requires computation at the local level. Which is to say, swarming is an emergent behavior from being able to self-organize. I would imagine most organisms have some kind of drive for attraction and repulsion, whether its mates or predators. Those two instincts can drive the computation for swarming. I'm kind of applying some semantic flourish here but avoidance is kind of vaguely high-level concept for computational behavior. It's definitely not as cut and dry as what I said but you can certainly program it that way if you have a high-level language. I'm not saying it's difficult to understand by the way, I'm saying it's hard to write out predict complex behavior from simple rules because it's hard to simulate those systems. Self-organization falls into like, computational reality theories that are way above my head but a good place to start is cellular automata's theories of complexities.
Shorter answer: almost everything living on this planet is still here because it exploits some kind of metaheuristic that increases its advantage from there being more of it. To massage your statement, I think almost everything exhibits some kind of swarm intelligence. I think it's philosophically simpler to argue the following procession of ideas that swarm intelligence is a metaheuristic algorithm, all life is capable of computing these algorithms, but then ask the question what the hell is life exactly?
I'm being handwavy here because I don't have a rigorous definition for life and therefore can't answer you directly. Some kind of computational substrate that mutates and replicates? That's von Neumann's definition I think. You could probably do without mutation! Death and radiation already provide those conditions for you. Again, I'm curious as to how we self-organize in the first place. The hard part isn't how do we know how to fit ourselves into a larger system, but how do we reduce ourselves to a single unit? What makes an individual organism at a higher-level of abstraction one cohesive thing when it's actually made of layers and layers of processes?
Do viruses swarm? If you follow what I just said, that's if viruses compute the algorithm. But do they compute at all if they're technically just data? Is information itself a computer? What are we but games within games within games?
Nature really is the best computer science teacher if you want to be brutally effective at being inefficient.
https://en.wikipedia.org/wiki/List_of_metaphor-based_metaheuristics https://en.wikipedia.org/wiki/List_of_metaphor-based_metaheu...
- KineticLensman 6y ago> I think almost everything exhibits some kind of swarm intelligence ... I'm being handwavy here because I don't have a rigorous definition for life Perhaps before getting too philosophical about defining life it would be useful to provide a rigorous definition of 'swarm'. Then we can debate whether specific organisms swarm or not, as opposed to just being alive and in groups of two or more individuals. For example, my intuition is that orangutans never swarm in the way that starlings or anchovies do.
- AnonHP 6y ago> Nature really is the best computer science teacher if you want to be brutally effective at being inefficient. And at the same time, it also creates enduring and resilient organisms that propagate and thrive over time scales we cannot really and truly comprehend.
- andygcook 6y agoBy the prose formation, this seems like a GPT-3 generated comment.