I have attempted to adapt my understanding of ‘Who am I’ from my spiritual studies/experiments with that of my AI studies (neural networks in particular)
Spirituality places the question “Who am I” and getting an answer to it as the purpose of our living here. It counters the empirical nature of science by saying that “according to science my intentions, thoughts, beliefs, behaviors, self, identities, narratives all change from day to day and therefore we have no fixed identity”. Our identity then, according to science is a temporary shelter we take refuge in order to protect our own interests/whims and to live. In an era of postmodernism, we change our identities faster than ever by changing our homes, jobs, interests, partners, loyalties, appearances, beliefs, behaviors etc. so much so that a person changes every six months and becomes a new avatar discarding his old self like an older version of the latest software update. It is impossible to associate a single narrative story to a person. He is not one, but many then. Spirituality and when I say spirituality, I am mostly referring to Adavaita Vendanta philosophy and it says that there is a fixed identity in you and there is a way you can know it and by and through knowing it, you rest in your consciousness center without the necessity to take these multiple avatars every day to please your new changing yourself and your changing impermanent world.
Now, moving to neural networks (I know I am attempting to tie two poles together here, the AI and spirituality, but if you bear with me you will have a revelation in the end) and covering the basics first. Neural networks/network of neurons in our brain, amongst many things magical they do, help us in understanding our world through the classification and association of this world into distinct physical objects, mental ideas etc. For example, neurons help us to classify a Rottweiler breed from a Husky breed. Not just that, if your brain loves Rottweiler dog and mine loves husky then it is because your neurons associate ‘like’ with ‘Rottweiler dog’ and my neurons associate ‘like’ with ‘Husky dog’. But how? This is because your brain has ‘biases/weights’ encoded in such a way in the neural networks that it fires ‘like’ only when Rottweiler dog is seen. Therefore, it is the biases/weights in our neural networks that dictate what we like/what we don’t like and what we classify. We can change the weights and change the liking. For example if I see a Pit Bull dog and like it, then I might start liking both or start disliking my earlier Husky dog depending on the bias/weights in the head. So, this all means, what we like/dislike/classify/observe/think/act depends on the weights and we are our bias/weights. Since all we do is like/dislike/classify/observe/think/act etc. we are our neural network weights/biases. Is it not? We can change the weights and become anything we wish to become. If you find your neighbor/partner ugly, you can change your weights to see them as beautiful. If you fear heights, then change the weights to become fearless. So we are our weights then, except that Advaita Vedanta has something different and more to offer.
Getting back to spirituality of Advaita Vedanta and the question of “Who am I”, it says that to know yourself you have to become nobody by giving up all the million avatars you take. When you become nobody, you know what this somebody you have become today. You will know “what is” by being “what is not”. Otherwise you will be moving from somebody/avatar to another somebody/avatar and these avatars will constantly show the world through your avatar. For example if you have donned the avatar of a zoo goer today, then you will see a rope lying in the distance and think of it as a snake, but if you have donned the avatar of washing clothes, then you will see a distant snake and perceive it as a rope to hang clothes. If you have donned a childlike avatar you will get attracted to every toy store around you and if you have donned a fashionista avatar you will get attracted to the next big thing in fashion. So by changing from one avatar to another we see the mirrors of those avatars in the world and nothing more. If you are in a happy avatar, the whole world is happy and if you are in a negative mood then the whole world is unhappy. So, Advaita Vedata says that by changing the avatars you will see the reflection of your avatars and not the truth. It says that to see the truth of ‘Who I am’, I have to discard all the avatars and become nobody!
In neural networks we see the inner workings. We change our weights and change our lives (like/dislike/classify/observe/think/act). For example, in a zoo goer avatar, our weights/biases were tuned to the animals and therefore we saw a rope and perceived it as a snake. In the miserable avatar our weights were biased towards misery and therefore we saw the whole world as full of misery. So what we see in this world is just the weights/biases in our head or is the world really full of misery? How do we know? What is reality? Advaita Vedanta says put all the weights to zero and become a person without bias. It says clean your entire neural network and put all the weights to zero to know who you really are. This is because only when there is no bias/weights can you see the whole truth here. This is where the inputs from the true world represent themselves without the bias in the world. With bias/weights present in our neurons, all we see is a reflection of the presence of the bias/weights. We can always change the bias and change the world. But, in the absence of biases we see only what is permanent and true here. This is the wisdom of Advaita Vedanta.
And finally, the answer to the question “Who am I” is not the different avatars we take, but that permanence, that absolute and that infinite bliss which we observe in the absence of bias. Once we see the permanence, the avatars are simply a game. I have been there, and it works, 100% !