to Will: Sure you can make that program. Any program that has no random number generator, will always run the same based on the same input, thats a core concept of computer science. (given same input, and same database state)
to Matt: For that I would need your definition of autonomy. >From Wiki: Autonomy: one who gives oneself his own law) means freedom from external authority. Within these contexts it refers to the capacity of a rational individual to make an informed, uncoerced decision. In robotics "autonomy means independence of control. This characterization implies that autonomy is a property of the relation between two agents, in the case of robotics, of the relations between the designer and the autonomous robot. Self-sufficiency, situatedness, learning or development, and evolution increase an agents degree of autonomy.", according to Rolf Pfeifer. -So this appears to be "an agent that is controlling itself" Now I would argue that pretty quickly an AGI that is given a control unit that does not directly answer to a human (IE google style query answer, or directly commanded bot) would have autonomy. Now there is a restriction in autonomy, given by the environment, and to a degree, always from other entities. Humans are believed to be autonomous. But we must act within the laws of physics and our environment. I choose what I am going to do next. These choices however are limited by my "internal programming" IE the limits and bounds of what I know how to do, and what I am able to do. An AGI that has the ability to choose its next action would be autonomous as well, though still has to act within bounds of its environment. These choices however are limited by its "internal programming" IE the limits and bounds of what it know how to do, and what it is able to do. I would go much FURTHER and say that a complex agent such as a race-car opponent simulator is an autonomous agent, in that within its world realm, it has free choice of what it is able to do, even though it must still act within the bounds of the environment it is in. So I would say that an AGI is as autonomous as a person is under those definitions. For the consciousness argument I would take the same route. Point to something that a conscious human can do, that an AGI could not. James Ratcliff William Pearson <[EMAIL PROTECTED]> wrote: On 04/06/07, Matt Mahoney wrote: > Suppose you build a human level AGI, and argue > that it is not autonomous no matter what it does, because it is > deterministically executing a program. > I suspect an AGI that executes one fixed unchangeable program is not physically possible. Will Pearson ----- This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?&; _______________________________________ James Ratcliff - http://falazar.com Looking for something... --------------------------------- Got a little couch potato? Check out fun summer activities for kids. ----- This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=231415&user_secret=e9e40a7e
