A House Cat Knows More Than The IBM Watson

Is the mind the same as the body? What is consciousness? Can machines have it?

Moderators: AMod, iMod

Post Reply
SteveKlinko
Posts: 799
Joined: Fri Apr 21, 2017 1:52 pm
Contact:

A House Cat Knows More Than The IBM Watson

Post by SteveKlinko »

I have tried to avoid topics related to Knowledge, Knowing, Understanding, Recognition, Thinking, and etc., because it seemed difficult to define these things properly. In spite of that, I would now like to make an attempt at defining and talking about Knowledge and Knowing. Let us narrow the scope of the broader term Knowledge and define Knowledge as stored Information. In Computer Systems, Information would need to be Coded into a form of Data that the Computer can process. A Computer must have all Information converted into Data that consists of patterns of Ones and Zeros. In Brain Systems (Human or Animal), Information would need to be converted into a form of Data that a Brain can process. A Brain must have all Information Coded into Data that consists of Chemical Changes, Connection Changes, and any other changeable aspects related to Neural Plasticity. The important thing to understand is that for Computers and for Brains the actual Information must be Coded into some Physical Configurations in some sort of Physical Memory. With a Computer, the Physical Memory is the Hard Drive, RAM, or any other storage Media that can store the necessary Ones and Zeros. With a Brain, the Physical Memory seems to be distributed throughout the whole Brain and is implemented in the Neural Plasticity Changes of the Brain.

I think we can explore Knowledge and Knowing by submitting a test question to a Brain and then to a Computer. Let the question be: What kind of animal Meows? Putting aside the difficulties of parsing the Sentence, we can view this as a Memory Association Data Access operation, and the Brain and the Computer will produce the answer: Cat. But now let's examine the differences in the underlying processes that occurred. First for the Brain. We can say that the word Animal and then the word Meow will Associate in the Brain to eventually fire some Neurons that indicate Cat. But there is no Experience of Knowing that the answer is Cat, until the next Processing stage where a Signal of Cat and Knowingness is received by the Conscious Mind. The Knowing is in the Conscious Mind. The Physical Brain Knows nothing. It is just a processor. Now for the Computer. The Computer can have an Associative Database that will retrieve the text string Cat for the given input of Animal and Meow. But the Computer does not have the next processing stage of Knowingness that is implemented in a Conscious Mind. So we can conclude that the Computer never Knows it computed the answer of Cat. With a Computer it is always only Processing. There is never any Knowing. This might seem obvious, but a lot of people seem to think that Computers are operating like Brains and that Computers can in some way Know things. There is nobody home in a Computer, like there is in a Brain.

Now let's talk about this in terms of the Inter Mind Model (IMM). We will ignore details of the Processing that had to occur in the Brain or in the Computer, that was needed to arrive at the answer of Cat. For the Brian there will be Neural Activity that fire Neurons for Cat. For the Computer there will be Computational Activity that will load the text string "Cat" into three bytes of RAM designated as the location for the answer. In both cases there has been some sort of Activity that indicates the answer is Cat. Up to this point the Process has been mechanistic and Mindless for the Brain and for the Computer. There is no Knowing that it is Cat yet.

From the operation of the IMM we know that the Inter Mind (IM) is intimately Connected to the Brain, which in IMM terminology is the Physical Mind (PM). The IM is constantly monitoring the PM and is able to interpret the Neural Activity as referencing the word Cat. The IM converts this Neural Activity to the word Cat and to a Feeling of Knowing. The IM then sends the word Cat and the Feeling of Knowing to the Conscious Mind (CM). The Conscious Experience of the word Cat might consist of some vague Auditory Experience of the word Cat and maybe some hazy Visual spelling of the word Cat. Other related associations of everything you have ever Experienced related to Cats can arise. But the important thing is that you Know the answer is Cat in your Conscious Mind. You also have the Feeling of Knowing it. There was no Knowing with merely the Neural Activity of the Brain. From this we can understand that the Computer will not ever be abe to Know it has calculated or processed anything. There is no Connection to an IM or to a CM with Computers. But that is not to say that a Computer can not be designed to include such a Connection. The Technology is just not advanced to the point where we know how to do this yet.

We can now Logically say, because a Brain can Know the Information that it Processes, the Brain can have Knowledge. But since the Computer does not Know the Information it Processes, we can say a Computer can not have Knowledge, even though there is much stored Information. If a Computer is Programmed to Display the word Cat on a screen then the fact that the word is Displayed still does not indicate Knowing for the Computer. A Human Brain would be needed to read the screen and do some further Neural Processing to Know that the Computer computed Cat as the answer. The Computer still does not Know anything.

Let's see what a Cat can Know and what the IBM Watson can Know. The IBM Watson has access to much Information but it can not Know anything. A Cat only has a little Information but it can Know this Information. A Cat therefore can know more than the IBM Watson because, whatever a Cat Knows, is going to be more than what the IBM Watson Knows, which is nothing. Without Consciousness there is no Knowing.
SteveKlinko
Posts: 799
Joined: Fri Apr 21, 2017 1:52 pm
Contact:

Re: A House Cat Knows More Than The IBM Watson

Post by SteveKlinko »

I would like to add the following before the last paragraph of the OP:
We can now talk about Learning as the storage of New Information. For a Brain, the Neural Plasticity of the Brain will need to be changed in such a way that the New Information is encoded. There is now a Configuration change inside the Brain, and we can say the Brain was Reconfigured with the New Information. For a Computer, some New Patterns of Ones and Zeros will need to be included in such a way that the New Information is encoded. There is now a Configuration change inside the Computer and we can say the Computer has been Reconfigured with the New Information. As described above, the Brain will be able to access this New Information and Know this New Information. The Information is not Learned until it is Known. It will be New Knowledge. If the Information does not become Known, it is just a Reconfiguration and nothing more. A Computer will be able to access the Information but does not Know the Information. So a Computer cannot Learn in the same sense as a Brain. A Computer can be Reconfigured, but it cannot Learn. A Neural Net can be Reconfigured, but it does not really Learn. We say the Neural Net Learned, mostly by convention and for convenience.
Dimebag
Posts: 520
Joined: Sun Mar 06, 2011 2:12 am

Re: A House Cat Knows More Than The IBM Watson

Post by Dimebag »

I’d also like to make a distinction, that is, between declarative, and non declarative memory.

Because knowledge is a function of memory, that means knowledge can be either explicit or implicit. To say it another way, knowledge can be consciously known, or unconsciously “used” for behaviour, or basically “procedural”.

Now I think forms of AI, say neural nets, can actually form procedural memory, and thus, can unconsciously use that knowledge for shaping automated behaviours.

But the crucial difference is, the neural net doesn’t “know” the information. It has access to it. When given an identification task, or pattern recognition task, it can distinguish between say a cat and a shag-pile rug. It can produce the word “cat” when given such an image to identify. But, does it “know” what a cat is? I think not. Because all it can do is tell the difference between a cat and a rug.

If I ask it a question, such as, “what has fur, four legs, a long tail, and meow’s” it cannot produce the answer. Thus it has no declarative knowledge of cats.

That would be a further layer which would need to interface with that image detection, and interact with a language layer, such that the implicit cat knowledge can be consciously and verbally accessed. If such a neural net were programmed and trained, we may start to approach something like declarative knowledge.

Would it be conscious? I doubt it.
SteveKlinko
Posts: 799
Joined: Fri Apr 21, 2017 1:52 pm
Contact:

Re: A House Cat Knows More Than The IBM Watson

Post by SteveKlinko »

Dimebag wrote: Sat Mar 27, 2021 11:38 pm I’d also like to make a distinction, that is, between declarative, and non declarative memory.

Because knowledge is a function of memory, that means knowledge can be either explicit or implicit. To say it another way, knowledge can be consciously known, or unconsciously “used” for behaviour, or basically “procedural”.

Now I think forms of AI, say neural nets, can actually form procedural memory, and thus, can unconsciously use that knowledge for shaping automated behaviours.

But the crucial difference is, the neural net doesn’t “know” the information. It has access to it. When given an identification task, or pattern recognition task, it can distinguish between say a cat and a shag-pile rug. It can produce the word “cat” when given such an image to identify. But, does it “know” what a cat is? I think not. Because all it can do is tell the difference between a cat and a rug.

If I ask it a question, such as, “what has fur, four legs, a long tail, and meow’s” it cannot produce the answer. Thus it has no declarative knowledge of cats.

That would be a further layer which would need to interface with that image detection, and interact with a language layer, such that the implicit cat knowledge can be consciously and verbally accessed. If such a neural net were programmed and trained, we may start to approach something like declarative knowledge.

Would it be conscious? I doubt it.
Very good. But I am sure it would not be Conscious.
SteveKlinko
Posts: 799
Joined: Fri Apr 21, 2017 1:52 pm
Contact:

Re: A House Cat Knows More Than The IBM Watson

Post by SteveKlinko »

Now I would like to say some things about Human Memory in general. For a long time I was convinced that the Memory of things was strictly a Physical Mind (PM) function, but recently it has seemed to me that the only thing the PM does is store the Neural Correlates of Memories and not the Actual Memories. The PM can fire Neural Activity for a particular Memory or Knowledge such as for a Cat, but where does the actual Memory of a Cat come from? There seems to be no Explanation for how Neurons firing could possibly mean Cat or any of the thousands and thousands of other pieces of Knowledge stored in our Memories. It seems Logical that the situation for Memories is similar to the situation for Conscious Sensory Experiences. There is no Explanation for how we Experience Redness from Neural Activity. But we have Logically deduced that there must be some sort of Inter Mind monitoring the Visual Cortex. When the Inter Mind detects that the Redness Neurons are firing it produces the Redness Experience for the Conscious Mind (CM). Along that same line of Logic, it is deduced that the Inter Mind must be monitoring the PM for Neural Activity related to all the Memories that we might have stored. When there is Neural Activity for something like Cat, the Inter Mind detects this Neural Activity and produces a Cat Experience for the CM. The PM provides the signal (Neural Activity) that triggers the actual Memory that is located in the Inter Mind.
Post Reply