AI 2023. Meet ChatGPT. - page 119

New comment
 
onceagain #:

And to answer such questions, I have a perfect example that demonstrates everything there is to know about Reason.....

Let's take a pure (artificially created,... without "interference" generated by the body) Mind... and we are going to teach it to play chess...
Let the person who is engaged in inputting information into this AI be a complete sucker in chess....

So, unlike that team of scientists, software engineers and chess masters who have spent mountains of labour to prepare Deep Blue for a fight with humans, this... pure artificial Intelligence will only have to explain the rules of the game... and describe the chessboard with the position of the pieces on it.

The difference between Deep Blue and the AI from my example is that my AI technology is able to build all that huge tree of knowledge of the art of chess by itself (!)... without outside participation... from the raw "material" given to it... ... from the original "material" of the information given to it, listed by me above.... Look,... there is no "mathematical representation" of OUR understanding of the essence in it,... as there is none in Nature itself (and chess, meanwhile, exists perfectly well without it,... this "mathematical representation").

We don't need to understand anything ourselves, when we deal with true Reason...
We give it the initial information in sufficient volume - it gives us the solution we have found.... That's all...

(don't believe me? I have an example that confirms it too)


good example, but there is one but. how will the "mastery" in chess be obtained? by a very long search and memorisation of combinations? or by a quick but "reasonable" development of the game strategy by this artificial mind? and if the result of the search will be obtained as quickly as by "reasonable" thinking? if the result is the same in quality and speed, what is the difference?)))

my point is that the speed of obtaining the result plays a role in the subjective feeling of "reasonableness" of the object of research. The Universe can be intelligent, it just takes a long time to think, for example.

Hawking and Penrose liked to talk and argue about these topics.
 
Andrey Dik #:

good example, but there is one but. how will the "mastery" in chess be obtained? by very long brute force and memorisation of combinations? or by quick but "reasonable" development of game strategy by this artificial mind? and if the result of brute force will be obtained as quickly as by "reasonable" thinking? if the result is the same in quality and speed, what is the difference?)))

my point is that the speed of obtaining the result plays a role in the subjective feeling of "reasonableness" of the object of research. The Universe can be reasonable, just thinking for a long time, for example.

Well, here's a second example...

Imagine yourself as the Count of Monte Cristo,...who has just "settled" in his "new" cell.
Let's say you know a new game,...called Chess. You know the rules of the game, but you're not a fan.

And let two (choose any one you like) people, known to you in real life as world chess champions, be placed in your cell.

And so, one "beautiful" day, pining from the boredom of the cell, you offer your neighbours for entertainment a new, for them, game - chess...
You told the rules,... scratched out a chess board on the tabletop,... made pieces out of bread crumbs,... and-ee-ee....

Can you imagine what you'd see as the Count of Monte Cristo, lying on the top bunk... and watching two fans of a new game and natural geniuses become masters of the game?

I am hardly mistaken in saying that you will see how from the scantiest source material of the rules of the game and the initial position on the board, in the "space" of the players' brains, without any influence from the outside,... simply as a result of performing all the required operations on the information, a huge amount of world-class chess knowledge grows.

So "... in what way will "grasmeistery" in chess be obtained?"...
Huh?...?

 
onceagain #:

Well here's a second example for you then...

Imagine yourself as the Count of Monte Cristo,...who has just "settled" into his "new" cell.
Let's say you know a new game,...called Chess. You know the rules of the game, but you're not a fan.

And let two (choose any one you like) people, known to you in real life as world chess champions, be placed in your cell.

And so, one "beautiful" day, pining from cell boredom, you offer your neighbours for entertainment a new game for them - chess...
You told the rules,... scratched out a chess board on the tabletop,... made pieces out of bread crumbs,... and... and...

Can you imagine what you'd see as the Count of Monte Cristo, lying on the top bunk... and watching two fans of a new game and natural geniuses become masters of it?...

I am hardly mistaken in saying that you will see how from the scantiest source material of the game rules and the initial position on the board, in the "space" of the players' brains, without any influence from the outside,... simply as a result of performing all the required operations on the information, a huge amount of world-class chess knowledge grows.

So "... in what way will the "grasmeistery" in chess be obtained?"...
Huh?...?


and, where's the intrigue?))) I'm telling you, there's no difference between "fast but stupid", between "slow and clever", if there's no difference between these two results. the difference is only in the time of obtaining the result!

Here is an example from me: if you run a random number generator, then in some time all known and unknown future works of art, science, beauty, life, genome, everything that is possible will appear. the only question is the time required for the solution, then any "reasonableness" is questionable.

If you are God, for whom 100500 billion trl bill bill kilo zigo mego fema mema trilo years flies by like a blink of an eye, then there is no point in worrying about creating an AI.... just one day, and very quickly for you, god, everything you want will appear by itself, because we know the paradox, in which when we strive to increase the entropy of matter, not only life, but also thinking forms inevitably appear, then for us, as very quickly dying it is understandable to want a quicker result. however, the essence remains the same - subjectively the difference in getting the result is only in speed, and nothing else.

We can talk about "mind" all we want, but we will not get any closer to understanding it. we, as mortals, simply do not have enough time for that.
 
one of the assumptions about the nature of the human brain is the theory of the quantum mechanism of neurons. from this point of view, our thinking is nothing but a probability field arising in the cortex of the gm.
in short, our brain is a probability generator with a selective mechanism for choosing variants. gpt on maximal.
 
or to put it even more simply, our brain is 2.5kg of fat.
 
Andrey Dik #:
or to put it even more simply, our brain is 2.5kg of fat.
That's a very optimistic weight estimate :)
 
Andrey Dik #:
one of the assumptions about the nature of the human brain is the theory of the quantum mechanism of neurons. from this point of view, our thinking is nothing but a probability field arising in the cortex of the gm.
In short, our brain is a probability generator with a selective mechanism for choosing variants. gpt at maximal.
Neurons work on exactly the same principle as other cells, but instead of dividing, they form new outgrowths that can attach to the outgrowths of other neurons. Just like other cells they exchange chemical elements, feed, form breakdown products and so on. Because of this feature, they can form new connections that weren't there before. The funny thing is that they are created as false connections, like false information, fantasy and so on, but these connections tend to be destroyed quickly, having no reinforcement in the external world. Of course, we do not take into account any deviations from the norm. And those connections, which are formed in the process of everyday life, have a great reinforcement, so by means of it a person adapts to the environment. And there is nothing surprising in this, because even plants, which do not have a brain, are also able to adapt perfectly to the environment, having a different set of cells. It is hard to say where in all this process one can find reason or consciousness or intelligence or generation of probabilities. Most likely these are just some generalised terms that have nothing to do with real processes, but help us to estimate the level of adaptability.

Here we can confuse ourselves, endowing ourselves with consciousness, mind and other things because of the fact that we are able to realise our thoughts, i.e. as if there is someone else inside us who understands and realises what we think about. This question is not studied enough, but if we look at it from the other side, we can feel that the process of awareness or self-reflection is accompanied by certain reactions in the body, that is, apparently, when we think, the brain periodically releases certain portions of hormones and impulses and tracks the body's reaction to these hormones and impulses. It turns out that the initial thought impulses return back to the brain in the form of the body's reaction to hormonal echoes and general changes in the body, as if tracking its work. And such signal-feedback recursion slowly fades, creating the illusion of an observer. Maybe we perceive the functioning of feedback in the organism as a kind of observer of the thinking process. This is a kind of deception or a plug, which cannot be determined intuitively, without studying the brain and the organism as a whole.

To summarise, consciousness can be called everything that has complex and feedback connections, and such connections are present in almost all objects of the material world. The differences between living and non-living consciousnesses are purely formal. Living ones are considered to be more complex in structure, therefore more conscious. That's why I don't understand all this fuss about AI, because in essence nothing special happens, except for the complication of connections.

You can go on to philosophise, why is there an increase in the complexity of connections? Well, the same reason why their destruction occurs, only in reverse: under the influence of external and internal factors :) There is no evolution in the sense of the mechanism of complication inherent in nature, these are random events, otherwise there would be no increase in entropy in the absence of external stimuli.
 

For me personally, two facts explain and put everything in its place:...

- first, that at the moment when you are reading these lines,... when you are aware of yourself... and what is happening - there is not a single point (!) in your body that would participate in this.... There is not a single neuron that is aware of anything with you,... even a tiny bit of your picture (!). Every (!),...I emphasise - every cell of your organism is dragging its own,...stupid,...aimless and meaningless existence (I mean the absence of the cell's own understanding of the purpose and meaning of its existence).... But, at the same time,...the whole "pile" of these cells as a whole,...realises and understands something else...

- and the second, that the most correct decision about the way of reaction in the current situation is strictly and unambiguously described by the laws of Nature itself. It is enough only to collect the necessary data and to carry out the required operations with them in order to get this decision and start using it.

From these facts it follows unambiguously that the Mind is not something holistic.... and very special,...but only an EFFECT,...the sum of elementary actions of each independent,...cell living its primitive life,...knowing nothing about the fact that it exists,...nor about the fact that it happens to "think".....

If so, this effect (Mind) can be realised on any suitable "hardware" for the simple reason that it is much easier to reproduce the essence of the "elementary actions" of a simple component than to come up with a design capable of producing the whole "effect" at once. The key to "Where to start?" is the meaning of the word "reasonable" - providing a positive result at the lowest possible cost.

 
onceagain #:

And to answer such questions, I have a perfect example that demonstrates everything there is to know about Reason.....

Let's take a pure (artificially created,... without "interference" generated by the body) Mind... and we are going to teach it to play chess...
Let the person who is engaged in inputting information into this AI be a complete sucker in chess....

So, unlike that team of scientists, software engineers and chess masters who have spent mountains of labour to prepare Deep Blue for a fight with humans, the real... pure artificial Intelligence will only have to explain the rules of the game... and describe the chessboard with the position of the pieces on it.

The difference between Deep Blue and the AI from my example is that my AI technology is able to build all that huge tree of knowledge of the art of chess by itself (!)... without outside participation... from the raw "material" given to it... ... from the original "material" of the information given to it, listed by me above.... Look,... there is no "mathematical representation" of OUR understanding of the essence in it,... as there is none in Nature itself (and chess, meanwhile, exists perfectly well without it,... this "mathematical representation").

We don't need to understand anything ourselves, when we deal with the true Mind...
We give it the initial information in a sufficient volume - it gives us the found solution.... That's all...

(don't believe me?... I have an example confirming this too).

Honestly, I was preparing for a different answer.

To the questions"How to reproduce complex interaction of different systems in changing conditions of the modelled environment? How to solve these problems without understanding the essence of things?", (taking into account the previously said that statistics "doesn't care" about the mysteries of the Mind), I would answer: "having collected a sufficient amount of data on the operation of the necessary systems in the chosen environment, we can repeat their interaction without understanding the causes of the driving processes. The same is true for the Mind.". Am I contradicting myself?

Your answer indicates that you do not consider the statistical approach as a way to circumvent the mysteries of the Mind, leading directly to the solution of its problems, and since we have not seen anything more universal than this approach, it is difficult for me to imagine the technology that you mean. I can hint that it has to do with genetic algorithms, optimisation,... and maybe something new. I won't guess.

But even assuming a universal statistical approach that doesn't involve digging into the mysteries of the core, we need to provide an unimaginable amount of data and resources to build the model. And the frustrating thing is that Reason is still more efficient. So the mysteries will have to be dug into. ))

 

With absolute ignorance of the environment and the laws of driving processes, their reproduction can only rely on a statistical model built on a huge amount of pre-collected data. The larger the environment and more complex the processes, the more data they produce, and the more time and resources it will take to collect, prepare and train a statistical model, which, like Reason, will predict the next stages of interaction of systems, and new events and states. Add to this the work of additional algorithms and you get the need to use a supercomputer.

The mind "reads" using some mysterious universality that allows to do without inconceivable volumes of data and dramatically save time on building models. How it does this is a mystery.

1...112113114115116117118119120121122123124125126...216
New comment