-- kourosh - 15 Dec 2008
Most of the existing definitions for an intelligent agent are around thinking/acting humanly/rationally. These definitions are not very scalable; they usually dismiss the fact these walking programs (human beings) have this type of intelligence, only because it's given them more chance to survive. So the intelligence is a meaningful concept in the context of life (and this life probably needs to be in a dynamic environment involving competition, selection, survival etc.) Therefore the definition of 'life' or 'being live' is important to know what kind of intelligence we are looking for. This definition, however, seems not to be scaleable. I can think of three properties of current definitions of life/intelligence which are not scaleable: type, time and dimension.
To be followed by lots of pizza