Marvin Minsky

Marvin Minsky regarded as the father of Artificial Intelligence (AI) is no more. He left this physical world at the age of 88 last Sunday. This is a great loss to the world. But his exemplary works, his lectures, his books and his quotes still make him alive in this world and it will be a great guiding force for researchers in AI, Cognitive Sciences and other related areas.

Although I have read about his work in my younger days, the first time that I came across more seriously on his work was during my seven years stay at Infosys somewhere in the starting of the year 2010.  That was the time I got interested to know more about the area of Emotion Modelling and Emotion Analysis. I together with my colleague (Nirmal, one of the author of this blog site) started to think about what is emotion. Even though, at first, we came across many papers talking about emotions and its negative implications for a human-being in this world we had a feeling that emotion is not that negative as proclaimed in the works that we had collected, rather it should have more positive implications. Unless otherwise it would not have survived the evolution.

So, why did we think that there is more positivity into it. The reason is simple to say but hard to realise. Just think about a human not having any emotions at all. People will say then that what is the big deal he won’t show any expressions, that’s all. Answer to this is you are partially right but you have not understood emotion completely. You are seeing only the behavioral aspects of emotions but there are two more aspects to it. One is the physiological reactions that happens inside the body and the other one, which is the main thing – many will forget or not realise – is the inner feeling. So, when we study about emotions we may have to look into all three aspects. Coming back to the example that we were talking about unless you have emotions is there a way to change the monotonous way of thinking, to prioritize the things you need to do, to know/be aware of the importance of a thing/or an entity? This is what made us to think about the positiveness of emotions. That was the time we came across two of the works done by the legends in this field: (1) Herbert Simon and (2) Marvin Minsky. Herbert Simon in his work said that

“emotions perform the critical important function of interrupting normal cognition when some unattended goals require attention and servicing”       [Herbert Simon 1967]

Minsky in his book on Emotion Machine stated that

“emotions, intuitions, and feelings are not distinct things, but different ways of thinking”

The confidence and the ecstatic feeling that we got when we came across these two statements gave us cannot be explained in mere words. This turned a new leaf in our work. Yes, of course that’s why we call Minsky a legend, right? There may be many many in this world who would have got a new leaf in their life like us. This turned in a new chapter in our thinking and now we are sure that if we need to model emotion then cognition should not be left. Rather, we have to work towards a holistic unified model not a model that has “holes”in it!

All the more, Nirmal sent a mail to Minsky about some of our thinking on emotion treated as human’s resource for problem solving. Surprisingly, he had time to reply to that and that too with lots of positive notes! Such is his humility to answer even patzers in the field like us! That’s why he is a stalwart, is it not?

In fact, every page in his book on Society of Mind deserves at least one PhD work!

I had collected some of his quotes from the Internet and prepared a video out of that. All credits to the original posters of the quotes. This is a very small tribute to him.

 

Advertisements

A Journey through the World of Space and Beyond

Today was an afternoon to cherish. There are some events that happen and you forget when time goes on but there are some events that happen and sticks to your mind, inspires you and makes you wonder and motivated to ask yourself what I want, what have I done and what I want to do for the country! I am sure when Padma Vibhushan Dr. K. Kasturirangan’s inspiring talk for Plinth  2016 @ LNMIIT came to an end, students and even some faculties like me present there would have felt the same and will live with that kind of motivation and inspiration to do something good for the country.

Basically his talk was on the ISRO’s mission from 1970’s to 2015, his Chairman’s period from 1994 to 2003, his MP of Rajya Sabha from 2003 to 2009 and his present stay at INAE.

His talk started from his years where he started applying for research fellowships. He was very candid to admit that he did not have a vision or aim like to go for space organisation and think of sending satellites. His main aim was to do research work on Cosmic X-Rays sources in the space. But the events from then on and due to advice and guidance from the great Vikram Sarabhai and Prof. Satish Dhawan’s guidance took him to this stage. This shows that you may not have a specific goal in your life to achieve but if you take the events happening in your life to your advantage and excel in what you are doing you can achieve greater things in the longer run. This is a great lesson to the students.

Then he talked about his first project of putting a balloon in the atmosphere to measure some radiations in the atmosphere. First time when they tried on an early morning at 4am it did not succeed since the electronics behaved bad. But when V. Sarabhai met them and came to know about the failure he encouraged them and motivated them. Then within a week they addressed the issue and succeeded in their mission. When V. Sarabhai met them again and came to know about the success he appreciated the team and advised them that since you failed for the first time you would have now learnt ten times more than what you would have learnt when you would have succeeded in the first time itself. What you got from here – if you are with the right person and work hard then the success is guaranteed. Failures makes you sad at that moment but when you think in the long term it will be a stepping stone for your success.

When Prof. Satish Dhawan was the Chairman of ISRO there was an instance where there was a problem in the camera hosted in a satellite. He said in an internal meeting that the camera seems to be dead but an engineer told him that based on his simulation tests it will be fine after one week. This became a challenge between them to which Prof. Satish Dhawan promised the engineer that if he loses he will give his one month salary to him. To which the engineer, in turn, also challenged saying if he looses he will give his one month salary to him. After one week when the camera was switched on, it started working as expected and Prof. Satish Dhawan took a ₹5 note and signed and gave it to the engineer. Engineer got surprised and asked him only ₹5, you promised one month salary?! Prof. Satish Dhawan said yes in fact this money is five times my salary of ₹1 which I get from ISRO. He added that his actual salary is from IISc, Bangalore! These fun happening brought cheers to people and brought ISRO people together and helped them to succeed.

When Dr. K. Kasturirangan was at helm in ISRO he made the environment such that anybody can question anybody in the technical aspects and he added that this made the projects success percentage to grow. His main achievement which he himself admitted as the first success of GSLV launch was not done without problems. First day of the launch the rocket did not start. But he said that was the first time they had designed an aborting program that will stop the vehicle to stop if some thing goes wrong in the four boosters that surrounds the rocket. When they researched why the rocket did not start they found out one of the booster was working only 85% of its capacity so the abort program rightly stopped the vehicle from launching or else it would have created a mayhem. So even in the failure they learnt that the aborting program is working correctly! This is what is all about in real life – you may fail but take good things from it and move on do not get disgruntled with the failures.

His idea of lunar missions and planetary missions when put forth to the then Prime Minister A. B. Vajpayee, he named it Chandrayaan 1. When asked why “1” there, Vajpayee told him that why to restrict to one mission you can have many more missions!

He also talked about his Rajya Sabha days where he played an important role in the Nuclear deal with USA.

Reflection on my class

It has been years since this blog got updated. Its time to get started again!

The series of blog on “Reflection on my class”  as the title signifies will be a self-reflection on my classes in Unconventional Models of Computing (UMC) that I am presently taking for 6th semester bachelor students in the Department of Computer Science and Engineering, LNMIIT, Jaipur.

First class on the first day of the year 2016! UMC as the name itself signifies the approach to do computing is unconventional. But first before going to describe about what is so unconventional about it I wanted to broaden their view about what is computer, computing, informatics and computer science. My emphasis was about this statement – computer science is not about computer but about computing! I gave some examples to pass on the message that nature also computes. For example, bio-organisms always want to survive and in order to achieve that it has various layers of protection very similar to our human immune systems. So, if nature computes can we use that naturally occurring computation somehow to do computation for solving our own computational problems using in vivo experiments? That was my question to the students. I told them this is what is our classes on UMC about – we are going to see how to use DNA strands, peptides, cells and its natural interactions to build a computational model that give rise to DNA computing, peptide computing and cellular computing, respectively.

My next idea was to make them understand with a bit more insight into how computing takes place and at the same time the problem biocomputing models faces. This I wanted to clarify even before explaining the advantages for using those models. Perhaps this can be taken as a disclaimer! For this two objectives, I took the example of our human computing evolution from using pebbles, abacus, calculators and computers. I told them using pebbles is computing. All the more when we use pebbles we represent it for some other physical object. We are finding a relation, in fact an 1-1 function. In an abacus, the same representation happens but there is a manual intervention (to move the beads) needed to do the computing – computing here basically is arithmetical operations. We have an algorithm but that is executed by a human. Now comes the calculator it also does (mainly) the same arithmetical operations (little more heavy) but once the input is given no need for any human intervention. Operations are performed automatically by the circuits inside. Now comes the basic silicon model computers where there is a stored program concept where we are offered some basic operations and we can write a program based on those basic operations and store it in the memory and can be executed anytime, anywhere and many times too. Here too, no manual intervention is needed. But in biocomputing models I showed this picture (taken from DNA Computing book)

Screen Shot 2016-01-02 at 2.53.45 PM

This picture presents us with many questions. Some essential questions to ask are

  1. Processing happens in test tubes. So, if two test tube solution is going to be poured into one who is going to do that?
  2. Who is going to find which test tube has to be selected and which test tube has to be left out?
  3. We may use a robot to do both the above steps but in that case who is going to instruct the robot?

So, in general biocomputing models needs manual intervention. As of now there is no stored program or automatic execution of program concepts. But it has its own advantages if not why are we going to study these models, isn’t it?  We will see this in the next series of posts.