From: Joo Hock
Date: 8/27/2007 10:49:20 AM
To: undisclosed-recipients:,
Subject: Bucky Gp - Can man ever design a computer .....
Dear Fullerenes,
Can humans ever invent a computer, that can think outside the box? This was the challenging question posed by Chia Siew in our discussion on Sunday at MacRitchie, as we were reading Bucky's "Utopia or Oblivion" Bucky said possibly never, or at least not for a long long time.
.
It seemed impossible, however Bucky also said: "Only the impossible happens" It was a very stimulating discussion, as Chia Siew maintained her position that one day it would possible for the computer to be cleverer, than man.
.
Hmmmm........
.
What do you think?
From: Michael
To: Joo Hock
Subject: Re: Bucky Gp - Can man ever design a computer .....
Dear Fullerenes
Conventional computers are single processor and are purely logic based. Multi-processors 'think' faster but only logically.
However, there is AI (Artificial Intelligence).
AI can be applied in hardware or in software.
In AI, there is an INPUT, a Rules Engine, and an Inference Engine. These three operate fairly independently.
Some conventional computers apply this. The difference is that AI operates with 'back-chaining'.
We all have experienced 'back-chaining' reasoning processes when we visit our doctors.
When the doctor sees us, he will listen to what symptoms we have and suspect that we have caught (say) flu. Then he will ask questions pertinent to flu. If most of the questions are right, then we have flu. This is back-chaining.
In forward chain reasoning, the doctor will academically have to take an infinite number of bio-data about you. E.g.. Height, blood samples, color or eye, temperature, blood pressure, what you ate for the day, ...etc., then conclude that you have (say) flu.
With back-chaining, the computer does its own reasoning. For those who are technical, it is basically performing a search in a knowledge tree. After the whole process, the computer will also learn from its mistakes/successes and update its knowledge tree.
Now the million dollar question - Can future computers think out of the box?
Well based on back-chaining reasoning and knowledge trees or neural networks (another way of storing information and traversing information) then probably not. This is because the computer cannot think outside the information tree, even though the knowledge has been learned by the computer itself.
To think out-of-the-box, the computer needs to be inspired.
To be in-spired, the computer needs to be "in spirit". Does a computer have a spirit?
In theosophy, we learn that all things have life, albeit some of the commonly called non-living things are mere lower life forms.
For instance, sand grains have life forms, otherwise the silica in it would not be able to help the computer to store information.
Taking this further, we also learn in theosophy that all objects in the Universe stores information. This is the Law of Archarsha, one of the 105 Universal Laws that governs the Universe. This universal record is called the Archarshic Records. Those who are in One with All, can access these Records and know ALL.
Psychics can go into a room and touch the walls and know the room's history of events. But this is still based on the psychic's ability and not the walls (lower life form).
So, I guess my answer would be no.
Besides, do we want or need a computer to think out-of-the-box. It will be good enough if we can get humans to do it.
Mic
Conventional computers are single processor and are purely logic based. Multi-processors 'think' faster but only logically.
However, there is AI (Artificial Intelligence).
AI can be applied in hardware or in software.
In AI, there is an INPUT, a Rules Engine, and an Inference Engine. These three operate fairly independently.
Some conventional computers apply this. The difference is that AI operates with 'back-chaining'.
We all have experienced 'back-chaining' reasoning processes when we visit our doctors.
When the doctor sees us, he will listen to what symptoms we have and suspect that we have caught (say) flu. Then he will ask questions pertinent to flu. If most of the questions are right, then we have flu. This is back-chaining.
In forward chain reasoning, the doctor will academically have to take an infinite number of bio-data about you. E.g.. Height, blood samples, color or eye, temperature, blood pressure, what you ate for the day, ...etc., then conclude that you have (say) flu.
With back-chaining, the computer does its own reasoning. For those who are technical, it is basically performing a search in a knowledge tree. After the whole process, the computer will also learn from its mistakes/successes and update its knowledge tree.
Now the million dollar question - Can future computers think out of the box?
Well based on back-chaining reasoning and knowledge trees or neural networks (another way of storing information and traversing information) then probably not. This is because the computer cannot think outside the information tree, even though the knowledge has been learned by the computer itself.
To think out-of-the-box, the computer needs to be inspired.
To be in-spired, the computer needs to be "in spirit". Does a computer have a spirit?
In theosophy, we learn that all things have life, albeit some of the commonly called non-living things are mere lower life forms.
For instance, sand grains have life forms, otherwise the silica in it would not be able to help the computer to store information.
Taking this further, we also learn in theosophy that all objects in the Universe stores information. This is the Law of Archarsha, one of the 105 Universal Laws that governs the Universe. This universal record is called the Archarshic Records. Those who are in One with All, can access these Records and know ALL.
Psychics can go into a room and touch the walls and know the room's history of events. But this is still based on the psychic's ability and not the walls (lower life form).
So, I guess my answer would be no.
Besides, do we want or need a computer to think out-of-the-box. It will be good enough if we can get humans to do it.
Mic
.
.
.
.jpg)
Posts
No comments:
Post a Comment