So, the next problem to contemplate is the difference between humans and computers from a **structural **point of view. Please bear with me, some of this is rather technical in nature. I also understand that you believe that the human mind is “more” than I will say - namely that there is a “soul”. That is fine, but whether there is “more” to humans or not is irrelevant. If this post can
substantiate that the simpler (materialistic) view is sufficient, then there is no need for the hypothesis of a “soul”.
Computers have hardware and software. The hardware is the processing unit, the memory and the peripherals or (name removed by moderator)ut/output (I/O) devices. The software is the operating system. The processing unit, the memory and the operating system are not necessarily distinct. There are
cellular computers, composed of hundreds or thousands of cells (of course the number of cells could be billions or trillions).
Each cell has a finite number of states. Each cell is connected to some other cells. Each cell communicates with its neighbors, and each cell changes its internal state based upon its previous state and the information received from its neighbors.
Surprisingly, there is no functional difference between an old-fashioned sequential computer and a cellular - parallel - one. Whatever problem one can solve, the other one can also solve. There is nothing new about this. Von Neumann proposed a neural computer and proved, that it can solve any problem that a sequential computer can,
and it can reproduce itself. E. F. Codd even gave much more elegant proof to the question.
So we know that computers can solve problems and they can reproduce. They can also
learn, there are many programs that evolve, change and learn, modify their own internal structures. They can also change themselves randomly - if given some peripherals that can detect light or sound.
The simplistic view that computers are fully predictable, that they cannot do anything that was not programmed into them is simply false. Not even the currently existing, extremely rudimentary computers - barely a few decades old.
They can even recognize faces, which is a very interesting problem. We have no idea how we recognize a face, and yet we do. The point is that it is not relevant that the recognition happens in a totally different manner. It is not necessary that the computer uses the same algorithm.
The next question is: can they generalize and conceptualize? Why not?
To generalize and conceptualize is to recognize certain features, and disregard others. In a sense they can already do that. The JPEG algorithm which stores pictures in a compressed format is something along those lines. It discards part of the information, and retains the rest. The result is so close to the original that the naked eye cannot detect the difference. In a sense it is a very rudimentary form of “conceptualization”. (Side note: severly autistic people cannot conceptualize. When someone mentions a “dog”, their mind will bring up all the dogs they have ever seen, and not just a “generic” dog as we do.)
Nothing I posted above is speculation. What comes now, is speculation, to a certain degree.
Can a computer ever pass the Turing-test?
To those who might not be familiar with it, here comes: The Turing test says that if one can conduct a sufficiently long conversation with “someone” and during the conversation this “someone” cannot be told apart from a human being, than this “someone” has consciousness.
The conversation must be conducted “deviously”. The same question should be asked, formulated differently to test if the respondent recognizes whether the differences are superficial or not. Anything is accepted to “trick” the other party.
Today, no program can pass the Turing-test, though some take quite a long time to “unmask”. There are many computer programs that act in a very complicated manner, some of them can even give good psychiatric “advice”.
I believe that eventually the computer will pass the Turing test.
What is needed?
- Memory. Check.
- Learning. Check.
- Ablilty to conceptualize. Sort of…
The proof of the pudding is that it is edible. As soon as the first computer will pass the Turing test, the concept of the soul will be obsolete.
Humans have body and mind. Part of the body is the brain. If you substitute the word “cell” with “neurons” the previous paragraphs you will get the description of the functioning of the brain.
The post above is
not a proof. It is merely a
substantiation of the hypothesis that the concept of the soul is not necessary to the idea of “consciousness”.