If you told it what it was to be sentient
if you gave it the ability to traverse semantic networks, combine knowledge, add to the network it could have "original thoughts"
If you gave it goals as to what was good art or bad art, or better yet allowed it to "learn" what art best suited its own current understanding of what was good or bad, what made sense and what didnt, what comforted its understanding of comfort, then it would create fine art same deal with fantasy art
of course i know what you mean;). But i'd say there wasnt any magic to any of the things you just said, rather, they are all just aparant high level affects of very low level and simple algorithms which everyone has built for themselves and that everyone follows
I'm afraid this argument is covered pretty well by searles chinease room;)
I ask you this, what is the difference between a well programmed machine and an intelligent one?
In fact could it not be argued that we as humans are machines that are particullarly well programmed and able to traverse the semantic web of our minds pretty well?
Ok a machine that can pass the turing test is a bit trivial, and as you put it could be covered pretty well by prolog...but...surley a sufficiently complex machine with enough prolog statments could do everything people could do...would it be intelligent then?
We could think about this experiments AI implications by seeing how it attempts to meet the turing test (immertion test). It has been argued that to meet the turing test an AI needs:
Natural Language Processing
Knowledge Representation
Automated Reasoning
Machine learning
Computer Vision
Robotics
So, lets see. This machine attempts natural language processing (loose, win, draw). It attempts to represent its gathered knowledge about the game. It attempts to automatically deduce (and thus reason) that if it has 2 cards it has either won, lost or drawn. It attempts to learn the rules of the game and it uses computer vision and CBIR to anaylse card drawing (again completely unlead)
Although this couldnt completely pass turing's emertion test, as it is just a simple game, we can see how it is trying to go in that direction
Note also the complete lack of robotics. I'd recommend to leeds university the possibility of attaching some giant mechanical claws with the ability to manipulate cards on a table (and possible kill all humans...who knows)
Slashdot Mirror
If you told it what it was to be sentient ;). But i'd say there wasnt any magic to any of the things you just said, rather, they are all just aparant high level affects of very low level and simple algorithms which everyone has built for themselves and that everyone follows
if you gave it the ability to traverse semantic networks, combine knowledge, add to the network it could have "original thoughts"
If you gave it goals as to what was good art or bad art, or better yet allowed it to "learn" what art best suited its own current understanding of what was good or bad, what made sense and what didnt, what comforted its understanding of comfort, then it would create fine art
same deal with fantasy art of course i know what you mean
I'm afraid this argument is covered pretty well by searles chinease room ;)
I ask you this, what is the difference between a well programmed machine and an intelligent one?
In fact could it not be argued that we as humans are machines that are particullarly well programmed and able to traverse the semantic web of our minds pretty well?
Ok a machine that can pass the turing test is a bit trivial, and as you put it could be covered pretty well by prolog...but...surley a sufficiently complex machine with enough prolog statments could do everything people could do...would it be intelligent then?
We could think about this experiments AI implications by seeing how it attempts to meet the turing test (immertion test). It has been argued that to meet the turing test an AI needs:
So, lets see. This machine attempts natural language processing (loose, win, draw). It attempts to represent its gathered knowledge about the game. It attempts to automatically deduce (and thus reason) that if it has 2 cards it has either won, lost or drawn. It attempts to learn the rules of the game and it uses computer vision and CBIR to anaylse card drawing (again completely unlead)
Although this couldnt completely pass turing's emertion test, as it is just a simple game, we can see how it is trying to go in that direction
Note also the complete lack of robotics. I'd recommend to leeds university the possibility of attaching some giant mechanical claws with the ability to manipulate cards on a table (and possible kill all humans...who knows)