Tuesday 24 May 2016

Roger Schank Blog

Roger Schank Blog 

Roger Schank in the latest article in his blog argues the AI is way more than just keyword matching and that an AI program should be able to exchange thoughts, hypotheses and solutions with other programs/humans.
 
He was commenting on the current state of AI research in the context of the massive attention media is giving to the news about a Georgia Tech professor announcing at the end of his course that one of his TAs was actually an “AI”. In fact, the “AI” he was referring to was nothing better than programs such as MARGIE, ELIZA and PARRY which were written in the 1970s and performed simple keyword matching. 

He predicts of an impending AI winter 2.0 due to the skewed perception media and, in turn, people have about the real potential of AI. He says there are important questions to consider in the field of AI such as:
  • Can we build machines that think, wonder, remember, feel and understand?
  • Is it enough if we continue to build such machines which do not perform any of the above actions but are still useful in one way or the other?

Tuesday 29 March 2016

Plato and Platypus

This article is a summary of the book "Plato and a Platypus Walk into a Bar" by Thomas Cathcart and Daniel Klein. The authors argue that philosophy and jokes are two sides of the same coin. Both of them amuse us, tickle our minds and makes us think. The book is an attempt to explain philosophy through jokes. The authors call this approach philogagging.

Metaphysics - Does the universe have a purpose? What are the characteristics that define an object? Do human beings have free will?
  1. Teleology - What is purpose of our lives?
  2. Essentialism - What are the essential qualities that define an object?
  3. Rationalism - We acquire knowledge through reasoning.
  4. Infinity and Eternity - Is the universe going to last till eternity?
  5. Determinism versus Free Will - Do we have free will or are all our actions predetermined?
  6. Process Philosophy - Evolving God!
  7. The Principle of Parsimony
Logic - How to perform reasoning?
  1. Aristotle's Law of Non-contradiction - A statement cannot be true and not true simultaneously.
  2. Illogical Reasoning - Wrong reasoning!
  3. Inductive Logic - Reasoning is performed based on observation and generalization.
  4. Falsifiability - In order for a statement to be valid there should be some possible circumstance when the statement can be proved to be false.
  5. Deductive Logic - Reasoning is performed based on rules.
  6. Argument from Analogy- Conclusions are drawn from analogous entities or situations.
  7. Post Hoc Ergo Propter Hoc Fallacy - An event X succeeded by another event Y does not necessarily mean X is the cause of Y.
  8. Monte Carlo Fallacy - Each trial is independent of the other.
  9. Circular Argument - One of the premises of an argument is the conclusion itself.
  10. Respect for Authority Fallacy - False dependence on the veracity of a higher authority than verifying the accuracy of a statement.
  11. Zeno's Paradox - Achilles and the Tortoise
  12. Logical and Semantic Paradoxes - This following statement is false. The preceding statement in true.
Epistemology - What does it mean to say I know something? How do we gain knowledge?
  1. Reason versus Revelation - Is knowledge attained through human reasoning or through revelation from God?
  2. Empiricism - Knowledge is attained through sense data
  3. German Idealism (Immanuel Kant)
  4. Philosophy of Mathematics - Analytic versus Synthetic statement, A priori versus A posteriori statement
  5. Pragmatism - Truth of a statement lies in its practical consequences.
  6. Phenomenology
Ethics - What is good? What is right? Is there any absolute standard to distinguish between right and wrong?
  1. Divine Revelation - Ethics is decided by God.
  2. Platonic Virtue - Ethics is decided by human wisdom.
  3. Utilitarianism - The end justifies the means.
  4. Golden Rule - Do unto others what you want others to do unto you (as long it can be considered a universal law).
  5. Will to Power
  6. Emotivism - Ethics is decided by our emotional state.
  7. Applied Ethics - Professional ethics
  8. Psychoanalysis - Ethics is decided by our unconscious being.
  9. Situation Ethics - Ethics is decided by the situation under consideration.
Philosophy of Religion - Is there a God?
  1. Deism and Historical Religions - "The Force" versus Creator/Clockmaker
  2. Belief in God - Theism (God exists), Atheism (God does not exist), Agnosticism (Current evidences cannot prove existence of God)
  3. Theological Distinctions - Jewish; Catholic, Protestant, Baptist, Methodist, Jehovah's Witness; Buddhism, Zen
  4. Airhead Philosophy

Philosophy of Language - How do we communicate?
  1. Ordinary Language Philosophy - "I promise" versus "I paint", "nothing" seems to be a thing, "I love you" and "I love ya" occur in different contexts, "I believe in God" can have different connotations,
  2. The Linguistic Nature of Proper Names - Bertrand Russel (short descriptions),  Saul Kripke (rigid demonstrators)
  3. The Philosophy of Vagueness

Social and Political Philosophy - Do we need laws? Is it possible to have an ideal state?
  1. The State of Nature
  2. Might equals Right
  3. Feminism
  4. Economic Philosophies
  5. Philosophy of Law
Relativity - What are the things that are relative and things that are absolute?
  1. Relativity of Truth
  2. Relativity of Time
  3. Relativity of World-views
  4. Relativity of Views
  5. Absolute Relativity
Existentialism

Tuesday 22 March 2016

My Political Views

I am a center-left social moderate.
Left: 1.46, Libertarian: 0.65

Political Spectrum Quiz

Edit (02-05-2020): After 4 years, I took the quiz again:

I am now a left moderate social libertarian.
Left: 5.45, Libertarian: 1.46

 

Wednesday 22 April 2015

A Thousand Splendid Suns, Khaled Hosseini

This is a fascinating book, which changed my impression about Afghanistan altogether. It tells the story of two women, Mariam and Laila, against the backdrop of various tragic events that happened in Afghanistan over a course of three decades. Hosseini displays exemplary narrative style by changing the perspective from one character to another with each chapter.

It all starts with the Soviet invasion in the late 1970s, followed by the civil wars between rival factions and finally, the Taliban. Kabul, which was once a vibrant and cosmopolitan city, gets totally shattered by the constant fights. Millions of Afghans flee to neighbouring countries. The Buddha statues of Bamiyan, which were destroyed by the Taliban, is one of the many atrocities that took place during that time.

Mariam is the character I liked the most in the novel. Her father was ashamed of the daughter since she was born outside marriage. Her mother so jealous of her that she feared her own daughter might enjoy the luxuries of life which she herself was denied. Her husband had contempt for her since she did not give birth to a boy. But still, Mariam never complained and led a selfless life. The final moments of her life, when she evaluates her own life and feels that she has someone to care for in her life is very emotional.

In short, 'A Thousand Splendid Suns' is a must-read.

Monday 21 July 2014

Areas of Research in Computer Science

This is a list of some of the areas of research in the field of computer science. Computer science is a rapidly changing area of research. It is difficult to predict the future because of this tremendous growth. The amount of digital data being generated is doubling every day. Thus, there are many interesting problems around us and computer science can be used effectively to solve many of those problems.


  1. High Performance Computing - improve system performance through techniques like using GPU's for computing
  2. Database – store, manage and process data
  3. Data Warehousing – maintaining huge database with data integrated from many sources, usually many databases
  4. Data Mining (or KDD) – process huge amount of data and come up with interesting patterns (like cluster analysis, anomaly detection, association rule mining) in the data.
  5. Artificial Intelligence - Artificial Intelligence is a field of computer science which deals with the study of how machines can be imparted with intelligence so that they can think and act on their own with a given data set. It involves knowledge representation, reasoning, planning,  natural language processing, machine learning, computer vision, robotics and strong AI, so that a computer mimics a human being.
  6. Computer Graphics - representation of image and video data
  7. Image Processing - analyze and manipulate image through techniques like compression, noise removal, background removal
  8. Computer Vision - extract information through techniques like pattern detection and recognition
  9. Artificial Neural Networks – reproduce the nervous system
  10. Bioinformatics – store, manage, process biological data
  11. Cryptography – protection of data
  12. Semantic Web – convert entire data in the WWW into a standard format which can be read by machines, because traditionally web is designed to be read by humans and not machines
  13. Multi agent system - a system in which intelligent agents communicate with each other in order to solve a problem which is difficult for a single agent to solve. http://www.cs.iastate.edu/~honavar/Papers/power-ieee04.pdf MAS involve ongoing interaction without ongoing communication.
  14. Theoretical Computer Science – theory of computation (automata theory or formal language theory, complexity theory and computability theory)
  15. Cluster Computing - It is similar to distributed computing, but here the computers are connected in a specific way, for example, a LAN. Whereas distributed computing spans huge geographical areas. There may be huge difference between the hardware configurations of cluster and distributed computing. Cluster is tightly coupled while the other is lightly coupled.
  16. Distributed Computing – software components located on a distributed system, which involves many computers on a common network through which they communicate with each other. Few aspects of DC are CORBA, Socket Programming, RPC, RMI. The users perceive it as a single node. As opposed to a centralized system, a distributed system has multiple points of control/failure. Common characteristics are resource sharing, concurrency, scalability, fault tolerance, transparency.
  17. Grid Computing – sharing of hardware resources, grid is where interaction is with the system as a whole and not with any node(s) in particular.
  18. Cloud Computing – sharing of services, store, manage, process data on remote servers, it involves many services like SaaS(Gmail), PaaS, AaaS, Iaas and etc.
  19. Parallel Computing – divide same problem into small problems which can be solved simultaneously
  20. Crowdsourcing - it is the process of obtaining needed services, ideas, or content by soliciting contributions from a large group of people, and especially from an online community, rather than from traditional employees or suppliers.

Friday 23 May 2014

Save Congress!

General Election 2014 results were declared last week. Though not even BJP had expected that they would win with such a clear majority, the poll debacle for the UPA was sure to come. UPA-II has failed the people of India. The scams, 'gates', policy paralysis, 'mute' MMS, RSVP gang all contributed immensely to this excruciating defeat. Modi was successful in creating his own space in minds of the people, such that even stalwarts like Nitish Kumar had to face the ire of the people, for breaking away from the NDA. While, some shrewd politicians, like Chandrababu Naidu, took advantage of the Modi wave by contriving an alliance with the NDA at the last minute, which ultimately put him back into the political equation in Andra Pradesh.

With a person like Rahul Gandhi at the helm, Congress has touched new low. Everybody knows it, but I am just saying it. RaGa has to resign. He may be the great-grandson/grandson/son of previous Prime Ministers of India. But that does not make him capable enough to lead the Grand Old Party, founded by the our freedom-fighters, which led India to the Independence and has seen India through each and every milestone in the history of Independent India. But now, Congress has dwindled to a party even without a National Party status.

RaGa needs an introspection. He needs to understand his place and space. I wonder why there is nobody in Congress to talk boldly against him. Till now, the allegations have only reached Rahul Gandhi's 'advisers' and has successfully avoided him. Who were these advisers? Would somebody in the Congress party care to name at least a few of them? Praful Patel even went to the extend of blaming MMS, the Accidental Prime Minister.

Assembly elections in states like Delhi, Maharashtra are fast approaching and another string of defeats is looming over the party. 'It's now or never' situation for the Congress. The solution? Let Priyanka Gandhi step in, as being clamoured by some Congressmen. Her credentials? Well, for one, she is the great-granddaughter/granddaughter/daughter of previous Prime Ministers of India. She even looks like Indira Gandhi. As for me, the only potential I found in her is that she is at least better than her 'chotta bhai'. While RaGa always talks nonsense, Priyanka talks some sense sometimes. Let Pappu roam around with chocolates in his mouth, clutching the hands of his mother.

Nitish Kumar took responsibility for the defeat of JD(U) in Bihar and resigned his post the next day. Let this be an inspiration to RaGa, and may he leave the control of Congress party to someone with good credentials, experience and passion.

A funny message I received in Watsapp says, ' Forget Tigers, Save Congress, Only 44 of them left.' Yes, it is time to save Congress.

Save Congress!

Thursday 22 May 2014

GATE CS/IT Notes

Please use this only as a reference. I would recommend all aspiring GATE candidates to create their own notes, which I am sure, would help you in the long run.

GATE CS/IT Notes

Please let me know of any error or typo you encounter in the notes. Since I am no Donald Knuth, it is not feasible for me to send you any checks or monetary rewards in that respect.

General Elections 2014

This was an election where social media played a greater role. All the candidates used all forms of social media like Twitter, Facebook, Google Hangouts, for connecting with the constituency. This was more than necessary because a reasonable portion of the electorate were under 25 years of age. 

Developments and policies took front in this election kicking aside all caste equations, thanks to the AAP effect. No wonder why Modi won with 'Bha-Ja-Pa' getting a clear majority in the Parliament. Policy paralysis, scams of UPA-II, ineffective top leadership of the Congress, Gujarat model all created the 'Modi wave' in the country. It was painful to see even successful ministers Nitish Kumar saw the wrath of the people for splitting from the NDA, leading to his resignation as Bihar CM.

The people of India were fed up with the UPA rule and Modi could successfully create a hope in the minds of the people for a better tomorrow. He reiterated the same in Twitter after the election results were declared, which within minutes became the most retweeted message. The Gujarat model was the perfect testimony to his ability and strength as a leader, which unfortunately was absent in 'The Accidental Prime Minister'. MMS is the best economist India has ever seen, thanks to abolishment of the License Raj and opening the Indian economy for foreign investments catapulting growth of India in the early 1990s. But he was a not good leader, and was mute on many occasions, when he was expected to speak out his mind. 

The attacks on the BJP, by Raga and party, as being detrimental to the communal harmony of India failed to create any edge in the elections. The Grand Old Party failed to understand the needs of the people. What India needs now is good business, jobs, growth, zero inflation and world-class education facilities. Modi was careful in being mum of issues related to RSS, Hindutva and 2002 riots and set his focus of policies and developments plans in his speeches.

Tuesday 20 May 2014

Post-GATE experience

My GATE credentials are GATE-CS/IT 2014, AIR 626, Score 713, Marks 51.95 and General category.
     
IIT Hyderabad, 06/05/2014 - Here, I went to attend the test/interview for Ph.D in Computer Science. It was a not-to-easy task to reach the IIT from Kacheguda railway station. It was, in fact, a multimodal transportation - local train, express train and share auto. IIT is currently in a temporary campus. Some departments have already shifted to the permanent campus, which is expected to be complete in an year or two. The primary reason I went there to interact with the candidates since most of them had already completed their M.Tech. Also, since this is Ph.D, I will get TA. Again, since my subsequent test/interview is in Madras, I thought I would get a head-start by attending test/interview somewhere else. Considering the proximity to Madras, Hyderabad was the perfect choice.

All information regarding the test pattern is given here - http://cse.iith.ac.in/phdadmission/phdadmission_info.html. The objective section was pretty easy. For the theoretical section, I had chosen TCS. It was very difficult for me and some of the questions were above B.Tech/GATE level. After the test I felt have very less chance to get shortlisted. But finally I was shortlisted for interview. It seems IITs are interested in direct Ph.D candidates and we got an edge because of that. My suggestion is that if your aim is to take a Ph.D in Computer Science, then consider taking the option of a direct Ph.D, instead of going for Masters and then trying for Ph.D. It saves you lot of time and energy. My interview didn't go well. I was not prepared about my research areas or interests. The interviewer asked about my B.Tech mini/main projects. There was one question from Algorithms - "What is the time-complexity of an algorithm, when nothing specifically is asked like best-case, average-case or worst-case." My answer was average-case time complexity because the other two cases are extreme cases and less likely to occur compared to the average-case. The result came after about a week and I was not selected. They had shortlisted about 25 for interview and then they selected 9 from them.

Next stop, IIT Madras, 12/05/2014 - Here, I had test/interview for M.S. (By Research) in Computer Science. This is research-oriented programme and is not a substitute for M.Tech, which is course-oriented. The details are given here - http://www.cse.iitm.ac.in/admissions. They had shortlisted about 380 candidates for written test. The test had both objective and subjective section. The paper was pretty easy. I was feeling confident, yet tensed, this time. Confident because I had cleared the same written test 6 months back during the Spring semester admission. Tensed because I wanted to prove myself that my clearing the test 6 months back was not a fluke as I could not clear the interview then. Finally, the results were announced and I was shortlisted among 45 others.

The interview was 'okay-okay' for me. The panel was friendly and asked me my favourite subject. I said Algorithms. The questions were pretty simple. I feel I took more time to arrive at the solution. Were they expecting a quick answer or were they simply looking at the method in which you arrive at the solution? No idea. The interview results are not yet announced. Usually they select about 15-20 candidates for each semester. So I have a 0.5 probability of getting selected.

Next halt, IIT Delhi, 15/05/2014-16/05/2014 - Here, I came for an interdisciplinary programme called VDTT. There were only interviews and no written test - http://vdtt.iitd.ac.in/Admissions.html. VDTT is jointly run by CS, EE department and CARE in IIT Delhi. The first round of interview was good and went on for about 40-45 minutes. The interviewer asked questions from almost all CS subjects. I could answer most of them, except for some in Digital Logic - "How to create a latch using a 2x1 multiplexer." The interviewer was very friendly to me and I asked him some general question regarding the VDTT programme, which he explained. I was shortlisted for the second round of interviews, which are conducted by R&D projects@IIT Delhi and sponsoring companies like Texas Instruments, Cypress Technologies. I had three interviews that day. They interviewer is only checking your aptitude towards the area he is working on. Be attentive to the project/industry presentations. I was too lazy that in one interview I was asked to explain the project. I had no clue and I was shown the door immediately, after which I felt depressed. Since the other two interviews went well, I was hoping of getting selected here.

The results came the next day and I was selected for an R&D project. TADA!!! Now I was really happy that I'm gonna study in an IIT. I had few more test/interviews in queue and I decided to go on and attend them as well.

Pit stop, IIT Bombay, 19/05/2014 - Here, I had test/interview for M.Tech in CS(Research Assistantship). They had shortlisted about 300 for written test. The selection process is available here - http://www.cse.iitb.ac.in/page33. The written test was good and it was both objective and subjective. Since this was IIT Bombay, for which there will be cut-throat competition and due to large number of candidates, I was not confident this time. But in about 3 hours, the results were announced and I was also one of the 60 candidates who were selected for interview. Next part was similar to that of the VDTT programme wherein you are interviewed by research projects within the CSE department. There were 10 projects and I gave preferences for four of them. There was no interview for my first preference and they declared selection would be on the basis of your score in the written test. I had one interview which did not go well. It was both HR and technical. Since I had 2.5 years working experience, I was asked about the nature of work I did, which I explained very well. After seeing IIT Bombay CSE department, the lively atmosphere, I felt I need to study here.

As of now, 20/05/2014, I am waiting for my interview results of IIT Madras and IIT Bombay. I have interview calls for SERC and CSA departments in IISc, Bangalore on 02/06/2014, 03/06/2014. I am planning to prepare and attend them as well.

The result of IIT Madras was declared on 27/05/2014 and I was one of the 12 students who were selected for admission. After about a week, IIT Bombay results were also announced, but my name was not there in the list.

Final destination, IISc, 02/06/2014 - I attended interviews in three departments, namely Electrical Engg., SERC and CSA. There was no written test in EE department. But, I cleared written tests in the other two. All interviews, except the SERC interview, were not good for me. All I want to tell wannabe IISc-ians is that if you have a good foundation in mathematics, then any IISc interview is just a piece of cake. When the results were declared after a week's time, I was not in the list. It did not come as a surprise for me because I had not done proper preparation for the interviews, which I now regret. Always try for the best. Ample opportunities are there in life. It is up to you as to how you make use of them.

Finally, I decided to join IIT Madras for MS (by research) programme in CSE department. Listen, my brethren, it is nothing but sheer determination and passion that leads you to the gateway of success. 

Friday 25 April 2014

Research Sector in India

Research sector in India needs a boost. We have top-class research institutes, like the IIT's, ISRO, DRDO, TIFR, but research in India lags behind that of many developed and developing nations. For example, BARC, which has some of the best researchers and scientists of India, produces only 3% of the power in India. This shows how much room exists for growth and improvement in research sector. With the burgeoning requirements of power, technology, communication, transport, trade and commerce sectors, research industry is of utmost importance.

The Government needs to formulate policies towards this front. After all, what the Government can do is set up a favourable environment for research. It can divert more funds for research projects.


The technical education sector in India is in a dilapidated condition. Apart from few top institutes like IIT's, IIIT's, all other educational institutions are under-performing. It is heart-breaking to see millions of students waste their precious time every year, without any glimmer of an idea as what to do with their life. To some extent, there is no motivation and inspiration for the students from the part of the faculty and peer groups. No wonder, 90% of them end up in some IT or marketing company, only to realize a few years later that the road-side workers and hawkers get more job satisfaction than they do. But, this realization dawns upon the poor guy at a time when he has reached a point of no return.

More students should be encouraged into joining academics and research. I think this is very much relevant in the current scenario. More students writing GATE, JEST, UGC(NET) and trying to get into top-class institutes, than previous years, shows that it is proceeding in a positive direction. A concern here is that the student assistantship is very less. I have seen many students who are really interested in India, but had to back off just because of the meagre stipend or institute assistantship they get every month. For an average M. Tech student, 8000 rupees is his only monthly earning for the next two or three years. Had he joined some corporate company, his earnings would easily have had jumped to millions in a few years time. So, it is important that students should engage in research without expecting only monetary gains.

It is a common misconception that you need to be extraordinarily brilliant and intelligent to do research. Fortunately this is not the case. It's the desire and passion to do original and relevant contribution which should drive you forward. Let me quote Harsha Bhogle, the famous cricket commentator, who once spoke, "Beyond a point, ability or talent is the most useless virtue to possess. It's what you do with that talent that really counts." He goes on to illustrate the point with an example of Sachin Tendulkar and Vinod Kamble, who displayed extreme dichotomy in their attitude towards cricket. Both of them started their careers under the same coach and were in the same team till they parted ways. While one of them went on to become a cricket legend world has ever seen, the other fell out of the race midway at the young age of 22 years. In short, attitude and passion should be the USP of a research scholar.


A country can never aspire to be self-sufficient without the aid of cutting-edge research. All developed nations set apart a significant portion of their budget for research and related activities. Research attracts investments and as for India, it can help in eradicating brain drain, which has been paralysing our economy the past decades, while other countries reap benefits. It is fact that a more than 30% doctors, engineers and successful entrepreneurs in US are of Indian origin. Why can't we bring them back? This is a question to be debated upon. To some extend, I blame the cheap politics, bureaucracy and unfavourable attitude here towards research, which is the reason why many academicians have left India for greener pastures in the West.

My vision about India is that India should become a breeding ground, more or like a mother hen, for the top-class research scholars in the world. Let us hope that this becomes a reality very soon. Let us all strive towards it.