So you STILL think that AI understands what it types??

[nemos]

Yes, but can anyone prove that a robot has no subjective experience?

Subjective experience can only be had by the subject. Is there any certainty that the AI represents the subject? How do you generally judge what the subject is?
I used to hear that the subject is the observer, and the object is the observed. It turns out that the cat is the subject, and he should have the subjective experience. Is AI watching us?

I assume that the subject should have its own self, that is, a person, in order to have some reference point against which to evaluate the observed. I, it, they there, and how it relates to me or someone else. By the way, making mistakes or lack of understanding does not at all indicate that the culprit is not the subject. It's just that he's probably not a very smart subject, or maybe, in this particular case, he's got something to "expand" his consciousness. Well, is it possible that there are some mental problems - schizophrenia, for example. How would you like a schizophrenic AI? Will AI need psychotherapists and psychologists in the future?
To make mistakes is a well-known human trait, because the only true research method is trial and error, all others are only auxiliary methods. You won't know if you're wrong until you try, well, or you can just believe in something, why not. How would you like a miracle-believing AI?

[Skepdick]

Subjective experience can only be had by the subject. Is there any certainty that the AI represents the subject? How do you generally judge what the subject is?
I used to hear that the subject is the observer, and the object is the observed. It turns out that the cat is the subject, and he should have the subjective experience. Is AI watching us?

I assume that the subject should have its own self, that is, a person, in order to have some reference point against which to evaluate the observed. I, it, they there, and how it relates to me or someone else. By the way, making mistakes or lack of understanding does not at all indicate that the culprit is not the subject. It's just that he's probably not a very smart subject, or maybe, in this particular case, he's got something to "expand" his consciousness. Well, is it possible that there are some mental problems - schizophrenia, for example. How would you like a schizophrenic AI? Will AI need psychotherapists and psychologists in the future?
To make mistakes is a well-known human trait, because the only true research method is trial and error, all others are only auxiliary methods. You won't know if you're wrong until you try, well, or you can just believe in something, why not. How would you like a faithful AI?

Distinctions distinctions distinctions. Always end up causing all the mess.

If I am observing myself observing myself am I the subject or the object? Either choice is a category error. Because the subjective/objective distinction is nonsense.

[nemos]

...If I am observing myself observing myself am I the subject or the object?..

One does not exclude the other, so there is no contradiction.
And categories do not exist outside of our minds, they are only needed by our minds and created by our minds to help us navigate the chaos.
Do you know how a work of art is created? We take a piece of chaos of a suitable size, remove all the excess from it, and what remains are the works of art. The ability to hear music in the surrounding noise, to see form and order (category) in the surrounding clutter.

AI may not, or will not, need categories, but if it wants to communicate with us it will have to do so in the form of categories.

[Skepdick]

One does not exclude the other, so there is no contradiction.

Really?

Doesn't the category of non-subjects contain objects?
Doesn't the category of non-objects contains subjects?

By definition.

If one doesn't exclude the other then it trivially follows that some non-subjects are necessarily subjects and some non-objects are necessarily objects....

[bahman]

There is no understanding without subjective experience.

Yes, but can anyone prove that a robot has no subjective experience?

Everything experiences but an AI does not have the coherent subjective experience that we call understanding. It simulates a neural network instead of being a neural network.

[nemos]

Really?

Really, really !!
Well, at least I am. And purely subjectively, the only subject in the world is me, you are an object, for me.
How do you think people learned anatomy if not by dissecting other people?

[Skepdick]

Really?

Really, really !!
Well, at least I am. And purely subjectively, the only subject in the world is me, you are an object, for me.
How do you think people learned anatomy if not by dissecting other people?

Why are you bringing me into your equation?

I am asking you what you are when you are busy observing yourself.

Are you the subject or object?

[nemos]

Are you the subject or object?

Oh, you caught me - I am the subject and also the object. But if that's how you're trying to prove the futility of categories, why don't you go ahead and try to do without categories at all.
My self - is the subject
My body, my soul, my organs, my will, ... are the objects of study of my self. In general, my self is also quite intriguing and worthy of dissection.
Well, why do you need to digest such obvious things so delicately?

[Skepdick]

Oh, you caught me - I am the subject and also the object. But if that's how you're trying to prove the futility of categories, why don't you go ahead and try to do without categories at all.

Sure, no problem. Categories are temporal and need not exist beyond the the moment they've served their purpose.

Categories are most useful when they are ephemeral.

But for some reason philosophers like to cast them in stone.

Well, why do you need to digest such obvious things so delicately?

Because the act of categorization is never performed in a vacuum. It serves some higher purpose.

[commonsense]

Yes, but can anyone prove that a robot has no subjective experience?

Subjective experience can only be had by the subject. Is there any certainty that the AI represents the subject? How do you generally judge what the subject is?
I used to hear that the subject is the observer, and the object is the observed. It turns out that the cat is the subject, and he should have the subjective experience. Is AI watching us?

I assume that the subject should have its own self, that is, a person, in order to have some reference point against which to evaluate the observed. I, it, they there, and how it relates to me or someone else. By the way, making mistakes or lack of understanding does not at all indicate that the culprit is not the subject. It's just that he's probably not a very smart subject, or maybe, in this particular case, he's got something to "expand" his consciousness. Well, is it possible that there are some mental problems - schizophrenia, for example. How would you like a schizophrenic AI? Will AI need psychotherapists and psychologists in the future?
To make mistakes is a well-known human trait, because the only true research method is trial and error, all others are only auxiliary methods. You won't know if you're wrong until you try, well, or you can just believe in something, why not. How would you like a faithful AI?

Yes, but you can’t know if an AI experiences. You can’t know that it doesn’t experience. It is the same as the other person problem.

[commonsense]

There is no understanding without subjective experience.

Yes, but can anyone prove that a robot has no subjective experience?

Everything experiences but an AI does not have the coherent subjective experience that we call understanding. It simulates a neural network instead of being a neural network.

So you say, but how do you know that a simulated network doesn’t result in experience?

[bahman]

Yes, but can anyone prove that a robot has no subjective experience?

Everything experiences but an AI does not have the coherent subjective experience that we call understanding. It simulates a neural network instead of being a neural network.

So you say, but how do you know that a simulated network doesn’t result in experience?

Doesn't lead to thought.

[nemos]

Yes, but you can’t know if an AI experiences. You can’t know that it doesn’t experience.

You can try to force it to say something that cannot be read or inferred, but only derived from its experience.

For example.
1. Do you have your own (subjective) opinion? ... - if not, then what kind of subjective experience are we talking about here?
2. How to distinguish good from bad? ...
3. I want you to tell me your (subjective) opinion about whether I'm good or bad, if you don't know, try to find out (keep in mind that I don't have to be honest). ...
4. ...

[commonsense]

Everything experiences but an AI does not have the coherent subjective experience that we call understanding. It simulates a neural network instead of being a neural network.

So you say, but how do you know that a simulated network doesn’t result in experience?

Doesn't lead to thought.

How can you know that? It’s the problem of others, isn’t it?

[commonsense]

Yes, but you can’t know if an AI experiences. You can’t know that it doesn’t experience.

You can try to force it to say something that cannot be read or inferred, but only derived from its experience.

For example.
1. Do you have your own (subjective) opinion? ... - if not, then what kind of subjective experience are we talking about here?
2. How to distinguish good from bad? ...
3. I want you to tell me your (subjective) opinion about whether I'm good or bad, if you don't know, try to find out (keep in mind that I don't have to be honest). ...
4. ...

Very good. Still, it's possible that an AI could search the net for an answer that “fits” the question, even if dishonestly.