Recently came across some new Stuff. Wrote a code that trains itself to identify people. Give few true images, meaning ‘This is how people look like’ and give few false images that doesn’t have people in it, meaning ‘This is how people doesn’t look like’.
You run the program, give it input, ask it to guess people in it, it finds few of them, it goes wrong too..Then for few times, you correct it saying, ‘You went wrong here. This isn’t a person. It looks like one but isn’t really a person’. Program understands that, corrects itself and doesn’t go wrong for such situations again.Eventually it becomes so perfect that it nearly finds people correctly. as it may go wrong, there is ‘score’ attribute tied to every of its prediction to tell more about its confidence in detection.
These type of algorithms that teach themselves and learn from past experiences are called ‘Machine learning Algorithms’ and this trained program is roughly called ‘Classifier’.So I wrote a classifier to classify people. I can train it for detecting animals, trees, cars….any object in the world.
This thing raises a question in my mind…Is our brain really a smart thing? Or its just a set of Classifiers trained to identify different things?
Of Course it does processing very fast but thats not a smartness!!! its possible for it to do because of that infinite memory to store and lots of RAM to process and the best Sensors in the world. But It couldn’t be counted under Smartness.
A new born baby is like a clean slate we say. May be thats why it doesn’t understand a single word spoken with it…and thats why there’s this tradition to still speak with babies so that it will grab more words and train it language classifier.
May be that’s why that baby observes so much to everything around…actually the process going on behind is all about training different classifiers in that baby !!! Classifier for language, people, trees, sounds, walking, senses…
By the time that baby grows, its classifiers for different things are trained enough to survive in the world and for the advanced stuff, we have education that fine tunes those classifiers…
What is this thing called ‘Experiences’ then? Its the situations, the things where that particular classifier being weak, went wrong and its result corrected us…Mistakes we make are nothing but our classifiers going wrong…may be thats why the choices we make, the options we have in life go wrong sometimes…
Is the classifiers score I talked about is the reason behind our Liking ? Is it that I like programming more than maths because from my childhood until this second, my classifier for writing programs is more trained than that of Maths ?
Or Is that why a baby feels more comfortable with mom than others? because in its initial days where its classifiers for identifying people is clean, mom is the one who remains always there to it ? and is it why we love our parents more than anything in the world?
Is it because of which even a parrot can talk if it is trained? (Don’t misunderstand its ability of understanding with ability of speaking. His brain has got little power.Its too much that it can at least talk).
Is it the result of classifiers because of which We think of RED when we say the word red? If it is so , can we false train a person to think of Green for RED?
If this all is a result classifiers trained for 100 different things and passing every situation through them makes us aware of the world around us and the situations in our life and eventually makes us a better person? and Intelligent people then are just the people with better scored classifiers, trained best with the study, experiences and situations ???
Is it because of which a person gone through a large heart attack of may be some brain injury can’t even identify people in his family ??? because he/she looses few of the best classifiers in his brain?
And If this all is true? Can we alter murderers , rapists, thieves, gangsters classifiers trained for such strange things into something better by training them properly rather than keeping them in the jail only (word ‘only’ is important here.I don’t meant to release them).
If growing is just a process of training the classifiers and processing the visuals from the eye fastly, brain has got huge memory and best sensors for that. That can’t be counted under smartness
and If this all is true, then can we train a tiger to eat grass rather than killing other animals ? it will disturb the habitat but it isn’t impossible !!!