Why are brains so much better at learning/improvising than computers?
In: 7
Instead of thinking of the brain as a single computer, remember that it is made up of almost 200 billion cells, which are themselves fairly competent; individual cells can hunt, flee predators, find mates, & even learn. Single celled animals are some of the most successful organisms on Earth. So imagine the brain is actually a composite of almost 200 BILLION individual computers, a single neuron can have over 15,000 connections with other cells. Some people estimate there may be over 600 trillion connections in the human brain.
We’ve really only just begun to learn about the brain, our understanding of it is exceptionally limited, and grossly over exaggerated.
It’s all about specificity vs. generalisation. Computers process all information as numbers and therefore base their memory and “thinking” on very rigid mathematical algorithms. This means that they are able to process a lot of information very quickly, and recall that information accurately, but it isn’t good at making connections because it is so precise that either two things are the same or their not the same.
The human brain is nothing like a computer. It processes a different information from your different senses and instead of thinking in maths it is based on concepts. This is a much less precise, but much more flexible system. A computer doesn’t know what a cat is in general. It can know that these specific things are cats and will recall exactly how that specific cat looked. Your brain, however, focuses on the broad strokes so it can take the characteristics it has seen in the different things called cat and tell whether something new it sees is a cat or not, but it will never be able to perfectly recall how a specific cat looked.
The latest advances in AI, neural networks, is actually trying to mimic how the human brain works.
Brains are sentient. A computer is not a brain. It follows commands in program. It can process information really fast but it cannot adapt or make adjustments that are outside of the commands or limitations of the software.
People *want* things, and have the capability of figuring out through experience how to get what they want.
Computers don’t want anything, they just execute instructions. If you make a very clever set of instructions, you can sort of simulate a system that “wants” to accomplish specific things, like “recognizing crosswalks” or “picking the most likely next word in this sentence,” but these are very narrow compared to human desires as we interact with our environment.
Brains have the remarkable ability to learn and improvise because they can adapt, make connections, and understand complex patterns in real-time, which computers still struggle to replicate accurately.