Will there ever be a Super Intelligence?
A bit about our brain
We look the human brain under a microscope many years ago and we noticed that the brain cells are somehow different from our other cells, we called them Neurons, we studied them, experimented on them, and in the recent years we figured out how they work in great details.
A neuron has multiple inputs (medically called Dendrite), processing cell (medically called Nucleus), information transporting wire (Axon) and multiple outputs (Axon Terminals).
Neurons rest at -70mv, once they receive a signal or a collection of signals with total voltage above -55mv, the cell will create a new signal and pass it to the outputs, and so on. (this is very simplified, since the process includes mechanical, analog, and digital parts, but the final result, is a signal processing)
There are multiple YouTube Videos that explain it in great details:
When you look at it, this should definitely remind you of something 😊 looks like microprocessor does it? Well, the processor cell is a bit more structured, we wired them a bit different, but in general it is a very similar idea, signal processing, but with less inputs and outputs.
A typical human brain has 86 billion Neurons (processing cells), interconnected with each other forming about 100 trillion connections, processing information at all time, we use all of them (Regardless of that Sci-Fi movies are saying)
Our consciousness is the result of all these electrical signals, yes, we are the operating system and main app running inside this brain, people came very close to realizing this thousands of years ago, and they called it soul, today we call it consciousness, us, the one that runs inside that brain.
We are very good at image processing, we have two eyes capturing large number of partially distorted images per second, black and white mainly, with a few colors in between (smaller number of color sensors), the picture is upside down, part of the picture is missing from each of the eyes due to the nerve location (no sensors there) and other issues, still, the eyes will capture that imperfect image, the brain will start processing it through multiple stages until we get a beautiful image.
But it is not only the image, it is everything inside it, we know what it is and how far it is, a car can have so many shapes and colors and we can see it from so many angles and we still know it is a car, impressive processing capabilities.
It was hard for us to write software to understand sounds and images, but once we understood how our brain works, we started trying to make our computers more like the brain, but it takes a lot of time and effort to change the processor, so we wrote another layer above it to simulate neurons, and we call them virtual neurons.
From that point on, we can assemble any shape of neurons we want, and things started to workout, our computers can now process images a bit better (not as good as us, still very basic, but we are at the right track), our computers can also process sound better, they can speak with a very human voice, we are at the beginning, but we know the way.
There is a limit to the numbers of virtual neurons we can put together, and to mitigate it we are adding more processors in grids, and slowly changing the processor architecture to work better with multiple inputs and multiple outputs, at the other side, our brain is biological, the neurons perform some mechanical functions as well, this makes them much slower than silicon based gates, way slower, but the number of connections between these neurons inside the brain help mitigate that.
We now have the tools, and we know how the parts work, but we do not have a full brain scan, it took nature hundreds of millions of years until it figured out the current design for our brain, there is no way we can simulate and recreate that in a computer in a reasonable time, so we really need a detailed brain scan, does not look realistic, but someone found a very cleaver way to do it, and today we have a full mouse brain scanned in 3D neuron by neuron, connection by connection.
So, AI is under development, it is coming, but it is many years away, not because a technical limit, this time it is due to human understanding limits, and human communication and exchange of ideas, our brains can process billions of pixels of image per second, but still, we can only read or write relatively limited amount of words per minute, and we have hard time explaining ideas to others.
At one point or another, many years from now, we will get 86 billion neurons in a computer (human brain), fully operational, many years before that will get an ant maybe 250,000 neurons, a frog latter, 16 million neurons, and one day a human.
At this point we should decide, is this AI (Artificial Intelligence) or just I (Intelligence), just because it is silicon based it does not mean it is not real (artificial/simulated) it will be as real as me and you, he or she will feel better and will not get tired.
But computer-based intelligence will be very short lived, once we perfect the brain slicing and scan technologies, and the computer virtual neurons, we can easily then move from the biological body to silicon-based brains, it does not need to be virtual reality, we can be in this reality, just using a computer as the brain.
But this will not stop, there is nothing stopping anyone from running evolution at skyrocketing speed, and at one point or another, someone who started with 86 billion neurons will evolve to 100 or 200 billion neurons.
What can so many Neurons do?
This is a very easy question, the smartest gorilla has 33 billion neurons, and we toughed it sign language and basic human communication, she can say she is hungry, sad, happy, or tired, and a few other things, but in general, the life of a gorilla is predictable in great details, from birth to end of life.
Now take something very simple to us like a conversation from our daily life and try to explain it to the smartest gorilla out there, she may catch that you are sad or hungry or basic things a 2 or 3-year-old human can process, but anything beyond that is impossible. That gorilla has 33 billion neurons, a very large number, but still, it is not enough to process anything that is created by 86 billion neurons, just not possible.
It is actually much more scary than that, if you stand alone next to the gorilla unarmed, the gorilla will truly believe that it is much more powerful than you, and no matter what you do you can’t hurt her, there is no way for that gorilla to process something like a sniper hiding 100 meters away, ready to fire at your signal, it does not have the brain to process that, if it decides to attack you, it can be terminated on the spot and it will not even understand how it happened.
Once you pass the 86 billion neurons, you are more than intelligence, you are now super intelligence, and if it happens that difference between 200 billion neurons and 86 billion neurons is the same as the difference between 86 billion neurons and 33 (gorilla), then people who reached super intelligence will have understanding that is so advance that we can’t even process.
And here is the disturbing part, if this super intelligence decides to act against normal humans (us) we will not even see it coming, nor be able to process it.
This is the part that I wanted to reach in this article, everything around us has limits, basic things like the speed of light is limited, the number of transistors we can put in an area is limited, the number of simultaneously active transistor at a time before things overheat and melt is limited, and this continues from the most simple to the most complex.
In other words, the total number of neurons that can be put together and still work as effective is limited as well (unfortunately I don’t think anyway has that number calculated accurately yet)
We also noticed in our image processing (image perception) that the more processing we add to an image (beyond a point), the more errors we get, look at any image for a longer period of time and you will start notice additional shapes in things, now process this at very large speed in a computer and you will get the wrong results. it looks like that perception processing has its limits as well.
We also noticed that the smarter you get, the less you believe in religions and gods, and the more you question existence, and at one point or another you reach questions like “why are we here” and “what is the point of all of life” but at the same time our brains are wired in a way that does not bother the majority of us, and we also have so much more good things going on, so we ignore it.
But does that also mean that 86 billion neurons are the limit before people start questioning the meaning of life? Or it is perhaps 90 billion, or 100 or ??? and if there is a limit to that as well, does that mean that once someone reaches x number of neurons will decide to terminate?
But who am I do ask that question to someone with 100 or 200 billion neurons, we will be to them like a horse in a farm, big things that bother us maybe will be so simple or irrelevant to them. 😊