Neo-cortex, as we know it, has been the final upgrade by the Universe for the technology called ‘brain’. This has been the ultimate upgrade, responsible for social interactions & other complex cognitive functions. Now what if we can replicate or prototype this in machines?
Since last few years, with the advent of the concept of AI, we have been trying hard for this. This has to be a crucial step forward if we believe in strong AI which will be self-sustainable & which will think for itself while also growing & evolving constantly.
The good news is we are already quite far in this. The first key step here would be to study the human brain itself. Surprisingly, we have started forming successful equations for understanding the behavior of neurons within our brains.
If we try to understand the neo-cortex technologically, it is made up of structures of the neurons built up in symmetric modules like the G5 processes in computers. With this new upgrade in our brains, we also started coiling our brain inside out so that more information could be stored & we could evolve to carry out more & more complex functions. As a matter of fact, the neo-cortex is still constantly evolving at a phenomenal rate.
There have been efforts from worldwide in the last 15 years to study the neo-cortex systematically in order to understand where & how our perception or our version of the reality comes from. A project pioneered by Henry Markram has already developed a brain in the form of a supercomputer. They studied each neocortical column by dissecting each column one by one. And there emerged beautiful patterns. Through this experiment, we now understand that the brain’s best kept secret is diversity. Each neuron is different even though it is made of the same fabric, like millions of different versions of the same species. And no neuron in my brain is similar to any neuron in yours. Also that it is important to study the interactions as neurons do not randomly communicate with every other neuron, they carefully choose partners to communicate with. But we can understand how different circuitries or arrangements of neurons with different neocortical columns could work. And we can design a neocortical circuitry & make it come alive using some mathematics & equations. We already have equations on how the firing neurons can be made into electrical generators. We also further have the mathematics to understand how neurons collect information & subsequently create bolts of lightning within our head. We now know the communication process of the neurons which happens through electric shocks & the chemicals which are then released by that. And we’ve even got the mathematics of this communication process as well.
This effectively means that you only need to know a few equations in order to simulate neurons & make them come alive. However, the diversity & the patterns for each neuron is so wide that it would take 10,000 laptops to be connected to a super-computer to do calculations & understand the patterns of one single neuron. Here is the first such supercomputer, The Blue Gene (10,000 laptops in a compact refrigerator size), developed in association with IBM which Henry Markram’s team uses to study the neo-cortex.
They have already been able to study different neocortical columns successfully. The Blue Gene produces different electrical signals when stimulated with different things & all new information is stored within the supercomputer in terms of these signals. Now the next step they are trying out is how to project these signals outwards into our perceptual space. This would be projection of reality as perceived by the computer. And if we can do that, we can actually enter in to any kind of reality created by the computer. The concept of multiple realities could advance to a really crazy dimension.
This theory still needs a lot of testing like any other theory of prototyping human brain. But having reached this far, we can say this with an almost certainty that it is not impossible to replicate the human brain.
And this complete study or research could be really beneficial for us in 5 major ways:
- To understand brain related disorders in the process. 2/7th of the world populace today is living with diagnosed brain disorders.
- Technology can provide us sample brains to experiment with. There wouldn’t be a need for animal killing in the name of experimenting.
- If we are to head towards strong AI, prototyping the brain has to be the first step.
- If we create AI using our own brain model, it would be easier for us to tune our brain with machine brain at the same frequency, creating a synergy/harmony between the two.
- We could just be able to switch to different realities aided by the computer brain.
The Universe took some 11 billion years to come up with a mechanism such as a brain. The game could be long & tiring. To be patient is the key if we already have the belief system in place!