Areas of Interest

bullet

Artificial Intelligence

bullet It is really charming to think that 'a machine can think'. It is now possible for a computer to have natural perception as of human.
bullet Pattern recognition for Eyes.
bullet Sound Synthesizer for Ears.
bullet Natural Language Processing for reading.
bullet ANN for Brain.
bullet It will be interesting to note that, a computer, if it is capable of processing natural language such as English, Tamil etc., it can learn by itself by reading scanned books or by surfing the Internet.
bullet Computers can compose music, listen to it to find out whether it is good or not etc., all this can happen.
bullet But critics of AI says that a computer can never achieve what is called Intelligence. For me it's enough if my computer greets me 'Hi Madhu', and says some enthusiastic words if I am dull.
bullet Mr. Edison, founder of Big Blue (IBM), said "Men must think, Machine must work", but their product 'Deep Blue' itself had changed that (Deep Blue is a chess master).
bullet Whatever may be the case, in near feature we are all going to speak with something which don't have life, mind, emotion and soul.

 

bullet

Computer Theory

bullet This is the next thing I often think of. All of our computers are theoretically equivalent to a Turing Machine (Next time if somebody ask you 'what computer you have?', your reply may be 'Turing Machine'). Why it is not possible to design an abstract machine powerful than 'Turing Machine'. I feel, here lies the answer to the question "What is Computing?".
bullet It is not that everything in this world is computable. Some problems don't have computational nature. I can't write an algorithm for loving my spouse. Some problems have computational nature but don't have computational solutions. (E.g. Give me an algorithm for finding out 425645565674126463 is a Prime Number or Not)
bullet It is necessary for us to find out things that can compute for us, cheaper and faster than what electronics do now. DNA Computing is a promising field in this direction. To find out such a computing element we need to answer the question "What is Computing?".
bullet Next we must find out a Language that forms a good interface between we, the humans and the computers. How hard it is for us to learn a new computer language and to express our ideas in terms of Algorithms!
bullet Similarly, it is too hard for a computer to understand our language. Because our language has ambiguous and context sensitive constructs. If my Dad says, "Madhu, Open up!!", it may be either my mouth or the door of my house. I resolve this using my common sense. But, how can a computer understand that without being intelligent.
bullet My computer (Wipro Voyager) can sort numbers, plays songs, show movies provided everything has been given to it before hand. But is it possible for my computer:
bullet To generate algorithms of its own.
bullet To correct and upgrade, update the code by itself (I often dream that how it will be if my computer switches on by itself and threatening me to throw back  the CD I have inserted).
bullet To have the feeling of existence as we have.
bullet To find a Ubiquitous language for both human and computer.
bullet I feel terrible to think about two entities that can communicate with out human acknowledgement or understanding. Human Genome Project is one such thing. Where humans rarely understood what that genomic codes mean. But a fully equipped computer can be able to trace the meaning of the codes. So this means, we can get the meaning of the codes only if a computer tells us (It will tell as long as it is an Idiot Box).

 

bullet

Computer Game Programming

bullet This is the next field that interests me. The way the games are programmed is  fascinating. This field interested me when I tried to write a 2D racer game. I felt the difficulty of representing the hindrances and changes that comes in track on the course of race. Then I happened to play Need For Speed, which inspired me a lot.
bullet 'Prince of Persia 3D' is the next game, where the objects are texture mapped using plain BMP files. I replaced this BMP files with my own and I have seen exciting results (Prince became one of my friends!!!).
bullet Next game I hacked is the age old 'Prince'. Me and Prasanna Venkatesh of Information Technology and Management, did that by changing the animation sequence files. When you run the game, you will go mad. When you order your Prince to jump, he will run!
bullet Once in my Fifth Semester, I wished to develop a small game around 'Chith Vihar'. It is where our department is housed. But modeling difficulty of Chith Vihar, massiveness of the  code needed, made me to feel that I can't finish it before my course ends. So I dropped it. Using tools such as
bullet Game Studio.
bullet Dark Basic.
bullet Quake 3 cloning tools and its source code.

we can design games faster than ever before. Open GL standards can be used, where we can have all that is needed to design a game and to control its action. Thing is that it needs some time and dedicated team work.

bullet AI based games are tough to design (At least I  feel like that). They need perfect understanding of how Game Search Trees work. Or else we can use Predicate Engines for designing strategic games. I have AI as a subject in my 7th Semester. I have an idea of designing a modular game that plays within itself and gathers experience, so that, it will be tough to get it to its knees.

Thanks for patiently reading my interests. If you too have interests in these fields, mail me about your thoughts and ideas. Any criticisms and contradictions are always welcome.

Back Page  Home Page Next Page

(c) Creativity Labs - 2003

Hosted by www.Geocities.ws

1