> Indeed, the chatbot is less real than the character, because the character is the product of another mind, while the chatbot's words are the product of complex mathematical operations conducted over a massive database of all the words humans have uttered, arranged by their frequency in relation to one another.
How is a fictional character the product of someone's mind, but a character generated from a massive database of words from other people's minds is not?
Believing a specific architecture is not capable of consciousness is a different matter. I am seeing statements that simply state things like "it is software or a machine and therefore cannot be conscious" without any further clarification. That clearly betrays a belief in souls.
Why? What's so strange about a human brain? It is a physical system of molecules running electrical and chemical signals. We already have direct proof that physical systems support consciousness. So why do you believe other physical systems can't have consciousness? Give me your definition of what consciousness is.
You don't have to know that. All you have to know is that you can feel and is conscious, and this consciousness is a result of your physical body..
Yes but you are not me, I am me. I don't know you are conscious or can "feel". Both you and an LLM present themselves using words and other external indications only. If you argue brain scans, then similar experiments have been done on LLMs too.
>but how do I know you are not faking the outputs of feeling?
You won't..but it does not matter.
>Where is your disproof that other physical systems can't hold consciousness?
I said no such thing. Just that there is no reason to think that LLMs are conscious. If we find some natural or artificial entity with some unknown and un-inspectable thing inside it, that behaves in an intelligent manner, I would not be sure that it is not conscious.
Why unknown? The human brain is known and inspectable, then you should say it is not conscious. You haven't yet shown why you think an LLM can't be conscious. You are just saying "just because" with no reason or parameters given.
So what is the reason to not think it can be conscious? Brains are physical systems, so are LLMs.
How is a human brain not inspectable? Tell me what part of it is beyond inspection? You have been making up absurd non sequiturs throughout this chat, ending with this beyond absurd statement now.
So are rocks. Are rocks conscious? Idol worshipers might have a point after all!
>Tell me what part of it is beyond inspection?
Oh, you can dissect a brain as much as you like. Is that what you mean by "inspection". I was thinking more like "How does it work". In that sense, we don't know a lot.
Rocks don't talk to me and think and act intelligently. Current LLM's do.
We don't know a lot about how chaotic systems work either in terms of predicting its outcome even if we know every individual step. The argument you gave for the brain equally applies to the LLMs, or any other complex system. So your premise is wrong.
Define consciousness right here, otherwise we have nothing else to discuss. You keep making up undefined crap and circular "it is it, because it is it" nonsense without proof. If you cannot immediately and clearly define consciousness in a testable clear manner right now, I will ignore and stop all conversation with you until you do.
Potentially past models could have been. None before came close to being this articulate and able to think and talk on our level before, and do things beyond us like difficult math problems.
You said "rocks don't talk to me and think and act intelligently." But a coma patient who is conscious would be unable to talk and act in an intelligent manner. That does not means that they are not conscious.
So consciousness does not require externally observable behavior. That was my point. So a rock could be conscious for all we know..
Also, If you are irritated by my "crap", why don't you stop replying? Are you paid or somehow obliged to correct every crap talker on the internet or something?
I am open to the concept of every entity having some measure of consciousness. At least it is internally consistent. As of yet though, if you want to go anything at all beyond "Only I exist", you will have to see that certain organizations of physical systems appear to have consciousness and certain organizations don't. You haven't yet given any definition still of what exactly you think consciousness is, and are trying to act special by thinking your own, probably not well formed, definition is the only right one.
It is a waste of time to talk to you as you keep refusing to define what you think consciousness is but keep talking as if its a settled matter that you know what is and isn't conscious. So correct, I see no further point in talking.
And I am sure that lolakutty will momentarily bestow the world with a consciousness test, solving a problem that has eluded millions of top thinking man hours over millennia of trying...
However, I feel rather certain, that deep down the proposed test will just be another "is it chemical goo?"
The test can only be done on oneself, and not on others. So I know I am conscious, and I am made up of a brain, among other things. So I know for a fact that this set of things that I am comprised of, can generate consciousness.
But then you are stuck being the only conscious entity...
To extended anywhere beyond the firm ground of "I know I am conscious", you need to step onto the progressively shaky ground of assumptions, which just amounts to "tests of consciousness"
i.e. I have a chemical goo brain that is conscious, therefore other chemical goo brains are conscious...
It is possible that you are the only conscious entity in the universe, and you have no tools whatsoever to disprove that. That knowledge is no help in determining the consciousness of anything besides yourself.
>Sure, if your architecture include hardware that is known to be capable of generation consciousness, the no one will say that.
Then circling back to your original comment, no architecture can include hardware that is known to be conscious, unless that architecture includes your personal brain, because lolakutty's brain is the only thing in the universe that is capable of generating consciousness...
Which is a totally meaningless statement. This is like im14andthisisdeep levels of philosophy here...the "Nothing is real therefore everything is real" kinda stoner riffing
He gave the classic "I feel" argument. The problem is I am not him. To me lolakutty and an LLM both can only present their feelings or state via external indications. Both he and an LLM can say they are conscious. Who if any of them is lying?
Nah, it "betrays" an awareness of something implicitly obvious: All the software/machines we might discuss are ones from ~2026, rather than ones from ~2226.
Similarly, "pets can't do my taxes" is a reasonable statement, because it is implicit that it doesn't include hundreds of years of genetic engineering and cyborg implants.
Pets can't speak human languages or do thinking on logic or natural logic. Current AI's can. Its a very bad example. People saying "current archs are limited" are much less than people outright claiming "software therefore not conscious". And I haven't yet seen any proof why the current AI's already don't have some consciousness or not.
> Pets can't speak [...] Current AI's can. It's a very bad example.
You've misunderstood the comparison analogy. It's showing two situations where implicit timeframes are important when people talk about what is "true", where the second one makes the importance extra-obvious.
"Pets doing taxes" could just as easily be swapped with "cold fusion generator in your garage" or "rocket-ship for personal commutes."
If you're trying to figure out how LLM weights are like the weights of a rocket payload, then you're going in the wrong direction.
I get what you were trying to say, I still disagree, its an open question and I feel existing LLMs already might possess some consciousness. I cannot prove it of course so I am not saying so definitively. You haven't given any disproof of the same either.
Those who think future systems could have consciousness would explicitly say so, indeed I have talked with a few who clarified as such. On the other hand I have seen a lot of statements that simply state without further clarification "software therefore not conscious".
delichon · · focus · HN ↗
How is a fictional character the product of someone's mind, but a character generated from a massive database of words from other people's minds is not?
WarmWash · · focus · HN ↗
hardbass · · focus · HN ↗
Terr_ · · focus · HN ↗
"Why can't you just admit that you think only tree-parsers have souls!?"
hardbass · · focus · HN ↗
lolakutty · · focus · HN ↗
Problem is that we don't know any such hardware..
hardbass · · focus · HN ↗
lolakutty · · focus · HN ↗
hardbass · · focus · HN ↗
lolakutty · · focus · HN ↗
It should have the ability to feel...
hardbass · · focus · HN ↗
lolakutty · · focus · HN ↗
hardbass · · focus · HN ↗
lolakutty · · focus · HN ↗
1. That you are conscious.
2. That you have a human body with a brain.
This implies that your body can harbor consciousness. That is what is special about your body in this context. And that answers your question.
hardbass · · focus · HN ↗
I am sure of my body, I am not sure of you. I can feel, but how do I know you are not faking the outputs of feeling?
1. Scooters can move from a place to a place. 2. That a scooter has a body with an engine.
Therefore scooters are special and cars don't exist.
lolakutty · · focus · HN ↗
You won't..but it does not matter.
>Where is your disproof that other physical systems can't hold consciousness?
I said no such thing. Just that there is no reason to think that LLMs are conscious. If we find some natural or artificial entity with some unknown and un-inspectable thing inside it, that behaves in an intelligent manner, I would not be sure that it is not conscious.
hardbass · · focus · HN ↗
lolakutty · · focus · HN ↗
Human brain is not "inspectable".
> LLM can't be conscious.
Because there is no reason to think so, and suggestion that would be like suggesting the existence of ghosts and goblins.
Extremely unscientific.
hardbass · · focus · HN ↗
How is a human brain not inspectable? Tell me what part of it is beyond inspection? You have been making up absurd non sequiturs throughout this chat, ending with this beyond absurd statement now.
lolakutty · · focus · HN ↗
So are rocks. Are rocks conscious? Idol worshipers might have a point after all!
>Tell me what part of it is beyond inspection?
Oh, you can dissect a brain as much as you like. Is that what you mean by "inspection". I was thinking more like "How does it work". In that sense, we don't know a lot.
hardbass · · focus · HN ↗
We don't know a lot about how chaotic systems work either in terms of predicting its outcome even if we know every individual step. The argument you gave for the brain equally applies to the LLMs, or any other complex system. So your premise is wrong.
lolakutty · · focus · HN ↗
These are not required for consciousness.
> Current LLM's do.
Computer chat bots are not a new thing. So were all the past ones conscious too?
hardbass · · focus · HN ↗
Define consciousness right here, otherwise we have nothing else to discuss. You keep making up undefined crap and circular "it is it, because it is it" nonsense without proof. If you cannot immediately and clearly define consciousness in a testable clear manner right now, I will ignore and stop all conversation with you until you do.
Potentially past models could have been. None before came close to being this articulate and able to think and talk on our level before, and do things beyond us like difficult math problems.
lolakutty · · focus · HN ↗
So consciousness does not require externally observable behavior. That was my point. So a rock could be conscious for all we know..
Also, If you are irritated by my "crap", why don't you stop replying? Are you paid or somehow obliged to correct every crap talker on the internet or something?
hardbass · · focus · HN ↗
lolakutty · · focus · HN ↗
I said no such thing..lol...If you are hanging to this, it is a waste of my time to talk to you.
hardbass · · focus · HN ↗
lolakutty · · focus · HN ↗
But at least, I gave my definition of consciousness, and at least there is one person who can test for it (oneself).
cindyllm · · focus · HN ↗
[dead]
levitatorius · · focus · HN ↗
WarmWash · · focus · HN ↗
However, I feel rather certain, that deep down the proposed test will just be another "is it chemical goo?"
lolakutty · · focus · HN ↗
WarmWash · · focus · HN ↗
To extended anywhere beyond the firm ground of "I know I am conscious", you need to step onto the progressively shaky ground of assumptions, which just amounts to "tests of consciousness"
i.e. I have a chemical goo brain that is conscious, therefore other chemical goo brains are conscious...
It is possible that you are the only conscious entity in the universe, and you have no tools whatsoever to disprove that. That knowledge is no help in determining the consciousness of anything besides yourself.
lolakutty · · focus · HN ↗
That is correct.
WarmWash · · focus · HN ↗
Then circling back to your original comment, no architecture can include hardware that is known to be conscious, unless that architecture includes your personal brain, because lolakutty's brain is the only thing in the universe that is capable of generating consciousness...
Which is a totally meaningless statement. This is like im14andthisisdeep levels of philosophy here...the "Nothing is real therefore everything is real" kinda stoner riffing
lolakutty · · focus · HN ↗
[dead]
hardbass · · focus · HN ↗
lolakutty · · focus · HN ↗
hardbass · · focus · HN ↗
hardbass · · focus · HN ↗
Terr_ · · focus · HN ↗
Similarly, "pets can't do my taxes" is a reasonable statement, because it is implicit that it doesn't include hundreds of years of genetic engineering and cyborg implants.
hardbass · · focus · HN ↗
Terr_ · · focus · HN ↗
You've misunderstood the comparison analogy. It's showing two situations where implicit timeframes are important when people talk about what is "true", where the second one makes the importance extra-obvious.
"Pets doing taxes" could just as easily be swapped with "cold fusion generator in your garage" or "rocket-ship for personal commutes."
If you're trying to figure out how LLM weights are like the weights of a rocket payload, then you're going in the wrong direction.
hardbass · · focus · HN ↗
Those who think future systems could have consciousness would explicitly say so, indeed I have talked with a few who clarified as such. On the other hand I have seen a lot of statements that simply state without further clarification "software therefore not conscious".