Re: [aima-talk] Illusory basis.
- No one has ever been able to give "definitions" of words and concepts such as "think", "mind", "perceive", etc. Thousands of books, and millions of papers have been written over the ages attempting to understand what these concepts might mean. And they "mean" different things to different people in different specialties, and in different ages.
To think that you could list a dictionary-style definition of these and then proceed is a mistake. As for "intelligent agents", the entire 1132 page AIMA book is an attempt to try to describe the concept with hundreds of thousands of words and many figures.
As for the "right thing", one way to understand more about that concept would be to read Russell's 1991 book, "Do the right thing".
In the hard sciences, the definition of "gene" has turned out to be extraordinarily complex - Nobel prizes has been awarded for people who make progress on our understanding of "gene" - no simple definition exists. In quantum mechanics, the definition of "momentum" is confounded by the uncertainty principle.
Your request for "definitions" should not lead to many answers or discussions, because people know that that is not the way to proceed. In your spare time, you might as well attempt to create brief and concise definitions for "love", "nation", "government", and of course, "universe".
- Bob FutrelleOn Thu, Apr 1, 2010 at 3:45 AM, zeldich.michael <zeldich.michael@...> wrote:
I has try to discuss my objections to the great book Artificial Intelligence, A Modern Approach, Second Edition, but did not receive any response from the authors.
Because of that I am decide to discuss them on that forum.
I will provide the chapter name and page number to make it easy to find out the original text.
Chapter 1, p1,2 1.1 What is AI?
"A system is rational if it does the "right thing," given what it knows."
That definition is not definitive. What is a "right thing"? What meaning have the phrase " given what it knows". It is not sufficient just to name something for a scientific purposes. The terms: "think", "mind", "thought processes", "decision-making", "problem solving", "learning", "perceive", "reason", "act", "mental faculties", "intelligence", "computational intelligence", "rationality", "intelligent agents" and "behavior" should be scientifically defined before it will be possible to use them for the scientific purposes.)
" knowledge representation to store what it knows or hears"
There are important assumptions behind that statement. We have to define the properties of an agent, which is capable to behave itself, interacting with an environment under control of a control unit. If that control unit is receiving information from a set of sensors, than it cannot directly derive the information about the causes in the World leading for occurrence of signals from the sensors.
That means, control unit cannot process the information about the World and cannot have any factual knowledge about it.
Best regards, Michael