PS: The Blog Header is designed by Ashish - The Future Emperor of the World

Thursday, April 17, 2008

An answer to my Family, Friends and Critcs

So all those who have been time and again asking me questions like, What is your problem? Why are you so irrational? Why can't you be more sensible? Why are you wasting your talent? Why are you not focused? Why do you behave in so stupid manner? Why can't you just act like others? Why can't you think in a normal way? Why do you pass such irrelevant remarks? And why do you ask dumb question which has no connection with the topic? What are your goals? What do you want? etc. And things like that, here is my answer.

First of all, I thank all those who have this irrational, unshakable, illogical belief that I have the potential to change the world which includes my parents, my dogs, and KG, besides others, sorry to disappoint you but I am well aware of my capabilities and my potential, I cannot!!! The majority who thinks that I am talented, or as Nadira* would put it "You try to come across as lout which you are not. " Well talent and lout are relative terms and you are entitled to have your own opinion. Ideas Differ!!!

Biologically I am classified as a "Homo Sapiens" and the reason why I am telling you this is that "Homo Sapiens" are considered "Intelligent". Being aware of my capabilities is to say I am aware of what is the scope of my interaction or environment. As a person I am supposed to act in a manner which will cause me to be most successful. A computer engineer thus defines myself as an "Intelligent agent"!!!

Now looking my problem from the view of AI and AI being the right term since we all acquire knowledge and then take decision hence our increase in intelligence is induced or "Artificial."

Now all those who think it is Greek and Latin let us define "Agent". An agent is anything that can be viewed as perceiving its environment through sensors and acting upon that environment through effectors. A human agent has eyes, ears, and other organs for sensors, and hands, legs, mouth, and other body parts for effectors.
The only thing that separates humans from Artificial agents is that we use both "bottom up" and "top down" approach to decide on our rational actions - Actions that will cause us to be most successful. However, as of now we have little or no amalgamation of these two approach in artificial agents.

To explain in a very lay man language these two approaches, If I ask you to learn 20 formulae of maths, you just get hold of the formulae and mug it up that is top down. All the things are just fed to the knowledge base ( memory ) of the agent. On the other hand bottom up approach deals with solving a jig-saw puzzle, nothing is known you try and find the best possible combination and learn by experience/interaction with environment.

Environment can be of different types:
Accessible vs. inaccessible.
If an agent’s sensory apparatus gives it access to the complete state of the environment,then we say that the environment is accessible to that agent. An environment is effectively accessible if the sensors detect all aspects that are relevant to the choice of action. An accessible environment is convenient because the agent need not maintain any internal state to keep track of the world.
Deterministic vs. non deterministic.
If the next state of the environment is completely determined by the current state and the actions selected by the agents, then we say the environment is deterministic. In principle, an agent need not worry about uncertainty in an accessible, deterministic environment. If the environment is inaccessible, however, then it may appear to be non deterministic. This is particularly true if the environment is complex, making it hard to keep track of all the inaccessible aspects. Thus, it is often better to think of an environment as deterministic or non deterministic from the point of view of the agent.
Episodic vs. non episodic.
In an episodic environment, the agent’s experience is divided into “episodes.” Each episode consists of the agent perceiving and then acting. The quality of its action depends just on the episode itself, because subsequent episodes do not depend on what actions occur in previous episodes. Episodic environments are much simpler because the agent does not need to think ahead.
Static vs. dynamic.
If the environment can change while an agent is deliberating, then we say the environment is dynamic for that agent; otherwise it is static. Static environments are easy to deal with because the agent need not keep looking at the world while it is deciding on an action,nor need it worry about the passage of time. If the environment does not change with the passage of time but the agent’s performance score does, then we say the environment is SEMI DYNAMIC
Discrete vs. continuous.
If there are a limited number of distinct, clearly defined percepts and actions we say that the environment is discrete. Chess is discrete—there are a fixed number of possible moves on each turn. Taxi driving is continuous—the speed and location of the taxi and the other vehicles sweep through a range of continuous values

Needless, to say we live in a environment that is inaccessible, non deterministic,non episodic, dynamic and continuous. In short, the most complex environment that can be possible. Not to mention it is also competitive where every agent works for its own benefit. Obviously, top down approach will not work in such an environment we need to have a bottom - up approach to be successful as amount of information or data to be fed to our memory will be too large to take correct decision every time.

However, our society is hopelessly addicted to the top-down approach which is though easier and very helpful but not a surety. I am learning agent, if you search wiki about it the information on learning agent, unfortunately it is not very helpful. As wiki describes learning agents: In some literature IAs are also referred to as autonomous intelligent agents, which means they act independently, and will learn and adapt to changing circumstances. This type of agent will be most successful in our environment. Note: Homo Sapiens is intelligent and I am a Homo Sapien. Hence by first order logic I am intelligent and thus qualify to be considered as a learning agent.

Structure of a Learning Agent

As seen fro the diagram, such agents have a performance standard as described by the creator like winning a chess game in minimum moves and/or losing as less pieces. In case of human the standard can be set as social status, academic performance, satisfaction level or happiness. It has a learning element which learns from the feedback given from the critic element which evaluates the performance. Like rating given to computer in a chess game is a critic for the software and feedback can be a certain move made in the game. In case of human our performance can be the feedback and critic can be our analysis, friend etc . We are never short of advise. A problem generator, can be a game scenario, or a condition as to move bishop before rook. In case of humans the problem generator can be brain which tells us to choose different road for exploring. I think, the structure is by now more or less self explainatory.

An interesting thing, to note over here is that, in case of learning agent the problem generator, can many times choose to under perform to gain more experience about the environment or in search of a better method. It may even overlook the tried and tested method in order to find a still more efficient way of problem solving resulting in a failure or under performance. The best thing that happens from this is that a deliberate under performance or a irrelevant move may eventually result in a better agent or may be an expert system. The keyword being to innovate.

As a learning agent I try to do the same! The only problem is that my memory is not always permanent and as a result my under performance or performance may not actually teach me something or it may not stay in my knowledge base ( memory) forever. However, I don't think this makes my behaviour is irrational.

If you see from my eyes, I am not weird the world is. I may be an expert system in the making ... Or as the world sees me now, a total failure. I prefer to take my chances!!!

PS: * Names have been changed to protect identity.
        The statements in italics are taken from Artificial Intelligence: A Modern Approach by Stuart Russell and Peter Norvig

No comments: