Printable View
I once met a christian that told me the dinosaur bones were put in the earth (by the devil) to test our faith..
I think that true artificial inteligence is realy hard to make..
on the other hand it is easy to make something that "looks" intelligent..
I can make a roibot that does stuff in a realy animal like way with realy limited stuff
(2 electric motors a couple of transistors, capacitors and some resistors, and 2 ldrs (light depending resistors)
It doesn't learn but acts realy animal like..
Ok, that tangent lost me... I'm not much of a believer, so I'm the wrong person to ask, but either way methinks genesis is more of an anology or similar.
Anyway, the first concepts came from outside stimulie, events occuring around us that are recorded as patterns, which then start being linked.
The same way every kid learns. We just got more efficient at communicating the patterns, so we don't have to start right at the beginning every time.
Peasure. how do we know pleasure. would a machine ever be able to pleasure my wife ?
Excuse my terminology here but I couldn't think of much else.
Pleasure is very basic... positive reinforcement from the released hormones. Basically a 'good' linkage is being stimulated. Associated with pro-survival modes of behaviour.
This thead is huge, so I don't know if the following had been already covered:
Many labs working on AI subject seems to use 2 main langage:
- PROLOG
- LISP
Their is 1 book often cited "Prolog Programming for Artificial Intelligence" written by Bratko
Networker - LISP has been touched upon but nothing on PROLOG yet.
I attempted PROLOG a few years ago but didn't have the capacity to understand it at the time, I should renew my efforts maybe now that I've started studying AI a little more.
As far as emotions, pleasure, etc go, what is the definition of intelligence? If someone can intelligently answer a question without getting emotionally involved and without getting pleasure from it then where do they come in? I don't think they're needed for something to have intelligence, they're simply another form of input, internal in this case, stemming from the external stimulation we experience.
I've written an AI program that ran off IRC, it simply associated phrases it heard with words in that sentence. While it could never convince any of its listeners that it was truly human, it did fool some of them for a short while. So just a thought, if you were to implement a weight system as has been suggested many times in this thread, it would seem possible to have a system that took sentences it heard and placed weights on the words that caused the sentence to be spoken in the first place, from this the program could respond to other sentences with the stored ones and seem somewhat intelligent. While this is far from self-awareness or true intelligence it's still a possibility for a chatterbot.
Continuing on the spoken language issue, if you read up on the theories that many linguists have about how we learn language you'll see that one school of thought is that we have the basics of language already programmed into us when we are born, this has actually been shown to be somewhat true, or at least appears to be true, when observing deaf children from different locations who have yet to learn any official sign language. Despite having never heard a word before, the gestures they tend to develop follow a similar pattern.
Anyone who is a fan of science fiction and AI/conciousness should read Frank Herbert's Destination Void which is the prelude to the Jesus Incident&Lazarus(sp?) Project.
-Maestr0
Mark, machines are pleasuring people's wives every day. Hell my neighbor has one that makes our lights dim everytime she uses it.
http://www.sexy-smut.com/masturbation/extreme/main.html
Again, "this thread is huge" so I am uncertain if this subject has been touched upon, but have you considered looking ino the foundation for AI, the so called founding concept by A. Turing, "The Imitation Game"?
Replicant
It is not the horror of war that troubles me, but the unseen horros of peace
well, to determine if a machine could feel pleasure, let's examine why we feel pleasure.
We are programed what is good and bad, some instinct, some learned. (hot stove- instinct, bugs- learned) our brain tells us if what we are interacting with is good or bad, the degree to which we think it is good determines our emotional reaction to it, which was mostly imatation of emotional reactions of our interactions as an infant. (what happens when you smile at a child?)
Sometimes some hormones are released too. well, I suppose a computer can't release hormones. however a computer can be programmed to detect environmental factors that could cause their release and imitate the result of their release in a human being. and a computer can be programmed to learn from interactions with humans, so I suppose that a machine could be programmed to feel pleasure, (or at least act as if it does, which I know isn't feeling pleasure, but hey, that's why it's called "artificial" intelligence)