We're talking about possible, uh, structures of Shivan cognition and thought that could explain how they act in FS1/FS2. We're asking: what kind of algorithm are they?
I mean 'algorithm' in a really simple sense. As a species or a civilization, how do they transform things they find into behavior? What's their map from 'this happened' to 'this is how we act?'
Let's talk a familiar algorithm: people!
We're each self-aware individuals. We can only directly experience our own life, so we don't want it to end. We have an array of instincts and a lot of learned behavior. We communicate by linking symbols to concepts, then arranging those symbols into language. We build social rules to teach each other to defer our own personal needs in favor of cooperation. When we encounter problems, we make a model of the problem in our heads, then we push the model around to see how we can apply our energy to solve it. This is called cognition. To build our models, we use both rules of thumb (called heuristics) and a volume of past experience, both direct (memory) and indirect (things we've been taught).
This is a really good system for us. We can pass information between each other with symbols. We can develop technology.
Most of us never really have to think about anything else. But there are other algorithms out there!
A rock, for instance, can accept energy and transform in response to the rules of physics and chemistry. It can 'remember' past experiences by changing its shape. But it can't do much else. It doesn't seek out energy and metabolize it. It can't assign symbols or build mental models of problems.
A simple worm can accept input and use rules of thumb, its instincts, to behave in a useful way. But it can't come up with new heuristics. To 'learn', it has to produce mutant offspring, some of whom may develop better instincts by chance. So the worm algorithm for problem-solving is evolutionary.
An ant colony has a set of genetic rules, called eusociality, which conspire to make each individual ant very interested in helping out its sisters. Each individual ant is quite stupid and limited. It has only basic rules, coded into its genes. But the interaction of many ants can produce collective behavior, like foragers calling for workers to plunder a dead beetle,or hundreds of workers linking together into a raft to cross a river. The ants even have very basic symbols: scent markers that can trigger behavior in other ants. So the ant algorithm is hive-based, and arguably a little cognitive, because it can take a problem and 'model' a solution out of collective behavior. But it's the hive doing the thinking, not the individual ant.
Another algorithm is the market. In a market, like the hive, individual actors have wants and resources. They spend resources to get what they want — trading things that are easy for them to find in order to get things that are scarce. Each one's acting locally, using only the information they have. But collectively, the market may figure things out that the individual actors don't know.
Another algorithm, seen a lot in machine learning these days, is the neural net. I wrote a bunch about neural nets but now I am deleting it because this post is too long already.
We're talking about what kind of algorithm the Shivans might be! Does an individual Shivan experience self-awareness? Does it possess 'qualia' (the internal experience of existing)? Do Shivans have the ability to create a model of a problem, solve it cognitively, and then move towards this optimal solution? Are Shivans aware that they exist, and that they're shooting spaceships? Do the Shivans care about local optimization the way we do — getting things done fast, with an economical use of resources and little risk of life?
