Eli5: the lens of your eye flips the image of what is happening in front of you and displays it on your retina, then your brain “flips” it again for you to perceive. What is the brain doing for that second flip?
In: 23
I don’t think it flips it at all. What would it even mean? There is no second projection. The image is projected on the retina, right? Like on a photographic material. And it’s not projected anymore so it doesn’t need to be “flipped”. Top pixels of the retina are just interpreted by the brain as the bottom of the original image. How does the brain know? You move your hand down, the retina pixels register the motion from lower ones to the upper ones. The mapping is original. It’s learned. The pixels can be assigned to random neurons, random “memory cells”, however, when you learn to see (and this is making sense of pixel data transmitted to the brain) you learn which pixels correspond to a particular area of the visible space around you.
The brain is a computer made of smaller computers. Its “camera” is a computer. Then you have a bigger computer which is the visual cortex. Then it’s a bigger computer that is your frontal lobes, a kind of a CPU, that gathers data from the smaller computers and makes a kind of sense of it. BTW, regular computers works similarly. They have a lot of chips that takes care of some details, they feed the results to other chips that work with the results.
Also, in some computers, like microcontrollers, you can map inputs and outputs to different physical signals. Let’s say you have some light sensors and some temperature sensors. It doesn’t matter where physically you connect the signals, you can map what goes where in the software.
Also, as I designed some machines like that – the machine can automatically detect what is connected to it. When we get certain data from certain inputs – we know what is connected where. It’s pretty convenient, because we don’t need to worry where we connect particular wires. We read the data and we know “oh, there it is”. ELI5, so let’s leave it why that automatic detection is rarely used and in machines we mostly depend on pre-determined connections. It doesn’t matter here, what’s important it’s possible, it’s pretty easy and that’s most probably how protein computers do that. It’s logical, since most of their programming and operation is achieved by the learning process.
You’ve probably heard about AI. It mimics how living brains work. It just learns how to connect the dots. You connect the dots wrong – you get “weaker” signal, you connect the dots right, you get stronger signal. If brain sees that connecting pixel data from the top retina pixels matches the data coming from the other senses, it’s the feedback it needs to build the correct spatial awareness.
Latest Answers