intelligent systems

I've been thinking too much of everything as a system.

An understandable system where I just have to trace everyone's actions from the beginning. But neural networks are too difficult to trace. It all happens too fast, and suddenly all you're left with is consequences.

I believed that if we're all to be part of an operational system, then all we have to do is perform the task we're given at the abstraction level we're at. For instance, given a situation where some wrong happens to me, the correct isolated action is to report it, and then all the relevant agents will respond and the action will be resolved because... why wouldn't it be? The system is aware of it. I, a single unit in this giant system, performed my part. I think I really believed that.

In an AI setting, this is very different. A single neuron can make a mistake, and it can not affect anything significantly, down the line. A problem with the very very end result is what matters, but no one really cares, or understands, the mess that happens in-between.

So, what are we? We're not solved systems, not a giant, magical, distributed machine that gets the job done and marches forward. But are we as cold as the neural network? Nothing matters but the end result?

Or maybe, we can be understood as a distributed system, but I misunderstood the purpose of the system as a whole.

Or maybe, someone forgot to, or neglected to, provide error handling.

Maybe this perfect distributed system with a unified known goal and great error handling can only exist in theory.

Biologically, as humans, we are unfortunately not controlled by systems, but instead intelligence. Systems may keep us alive, but intelligence is at the helm. And this intelligence we have is for sure better at doing what artificial intelligence has proven to be better at - making decisions without knowing why or how those decisions came to be. So, how can we be a system, if we are controlled by intelligence? We can't. Not everything can make sense to every unit of intelligence, every human. There is too much to process. If it could've been done, this perfect being would tell us exactly what to do, and we would be foolish to not follow their suggestions, this fictional being that understands multiple beings' intelligences in one being. So, it doesn't exist.

Do not simplify humans to systems.

Do not simplify humans to systems.

Do not simplify humans to systems.