We should not judge people by their peak of excellence; but by the distance they have traveled from the point where they started.

Thursday, July 31, 2014

Machine learning in the personal life ....

In Computer Science Machine learning is a field where based on past training set data algorithms self tune themselves to predict the future possibilities when similar conditions as already seen in the past arrive. Machine learning gets used at lot of places in the current Internet dominated world. For e.g. the ads that you see next to your Google search results are presented by monitoring your past search attempts and predicting what are the most relevant category of ads that you might want to see in future. There are many such examples. 

Common sense as we say in personal life is a lot related to Machine Learning kind of experience. However, machine learning works on some specific set of rules, unlike common sense which many times could be based on ambiguous statements which can not be correctly comprehended. That is one of the main reasons Computers / Robots are very good at doing highly specialized tasks but are very bad at common sense 
(or as we call it Artificial Intelligence) related tasks. 

Now coming back to the personal life. Interactions in social life in a new culture are many times based on machine learning kind of rules. We meet people. There are different phases our relationships with new people go through. We like some people while we dislike some. Those whom we like we continue to pursue, and as we move towards more intricate relationships where we start interacting more and more with these people, we start forming our own experiences, and patterns about how they behave, whether they really mean what they say and have the ability to stick to their words, or whether they say certain things just to please others always. An aware person tries to define patterns for these new relationships that are under development. 

Many times emotions skew the way people react, their needs to be in presence of other people, their communication ability etc. However, still one can form a pattern of what the category a person could be put into roughly. Over a certain period these patterns emerge very strongly and that is when we understand the true nature of a person, as we have gathered enough sample points to say with a certain level of confidence that this person is going to behave in this manner in future based on what I have seen him / her doing in the past in these different stages. 

Males and females show very typical and different patterns in this respect and on the basis of that one can really say they are different species altogether with different needs. 

What is attachment? An aware person is like the machine learning algorithm he knows what state he is in and what needs to be done to get to the next stage in trying to understand the other person and define the relation in the correct manner on the basis of what has been seen so far. An attached person is the one who is completely blind to the process of learning and is acting on the whims of his emotions without seeing the patterns. 

Why do westerners go through so many relationship break-ups and how can they handle so much emotional turmoil because in order to be in a sincere relation one has to trust the other person. In my opinion most of these relationships go through these machine learning kind of phases, where people are constantly trying to predict what the next stage would be based on past patterns. I think after a while the relations become an intricate machine learning kind of pattern matching thing, where people just do not care and move on if patterns do not match. Emotions run dry. 

A near perfect relation in this sense is then the perfect machine learning algorithm which can so finely predict the outcomes of people's emotions based on the already accumulated learning, and the trust that accompanies it. All other relations are like those imperfect machine learning algorithms which are flawed. 

Some random thoughts as I try to co-relate the Computer Science technology with the daily life ......

Saturday, July 26, 2014

Internet's own boy ...

I came across him when he committed suicide last year at the age of 26 years, while fighting the Computer Fraud related felony charges pressed by US government for downloading academic articles from JStor, a journal archival system. 

Then I read more about him. Tried understanding his work. 

An ardent Internet activist, one of the creators of RSS feed specifications, one of the Reddit founders, online campaigner and many more achievements in a short span of 26 years. 

He believed in the power of the Internet to bring an information revolutions for a knowledgeable society, to make this world a better place. His fight got more political during later years.

He was being charged for downloading copyrighted academic material from JStor, an academic journal from MIT campus. Jstor dropped the case, MIT took a neutral stance, however the government wanted to prove a point and make Aaron an example to deter future incidences of this kind, so they charged 13 charges under a Computer Fraud related act. After fighting it for around 2 years under a constant shadow of prosecution young Aaron finally succumbed. 

His death reviewed the debate on the misuse of such outdated laws and started the change to revive them. 

The documentary is a nice watch to get some understanding of the story overall. 

Aaron, only if the governments behave in a more sensible way to pave a way for a more transparent functioning of knowledge sharing, to build better societies, your soul would rest in peace...till then the fight would go on ....

Thank you ....

Thursday, July 3, 2014

How much smart is enough smart ....Is smartness contextual?

Are PhD's bound to be really smart people? I struggle with this question many a times. Largely because of the type of activities that I see many PhD's getting into.

A PhD is supposed to give a person exclusive depth of knowledge in a particular area. A PhD is supposed to bring out new knowledge and expand the horizons of already known knowledge.

However, how about common sense and general knowledge? Should a PhD who is so well equipped in a single subject not have interest in other subjects? For example lets consider a Science PhD student. The advantage of being a scientific PhD student is other areas of studies such as the Social Science, Humanities could be understood by these students, unlike the other way round. So a Science PhD student can understand what social science states by reading relevant books, but a social science student can not understand lets say what a Computer Science PhD subject states. If this is the case, do PhDs with Science background really take interest in other studies and try to have an impact in the general fields like common problems that many societies face? I know some Computer Science PhDs getting interested in socio-economic problems and trying to make an impact, but they are a very small minority.

A PhD in any subject is supposed to give the person the training to do research .  Research is supposed to be driven by curiosity and a drive to understand the answers to unknown problems. However, how many PhDs try to take it beyond their subject matter?

For example. as a PhD student one often rides bicycles. Some of these bicycles are really really crappy, so much that the amount of effort put in to go from place A to place B is enormous. Only if the student pays a little attention to get it in better mode, the outcome would be by and large much better. However, many times I have observed total lack of ignorance to such simple things. So the PhD student might be doing a PhD in optimization theory and publishing papers in related areas, but he might not apply it to solve his daily optimization problem of making a better cycle. Is that smart?

I have always come across these kind of examples. What kind of explanation could be given to such kind of a behavior? Is it that these things do not matter, as these people are so engrossed in their bigger vision kind of things. Really? I have no rational explanation.

This brings me back to my main question how smart is really smart. And how should smartness be quantified.