• Quick note - the problem with Youtube videos not embedding on the forum appears to have been fixed, thanks to ZiprHead. If you do still see problems let me know.

ChatGPT

arthwollipot said:
There - was that so hard?
Click to expand...
****! I've been saying the same thing in multiple posts if you would just think about what you are reading.

arthwollipot said:
Precise calculations are not specifically programmed in. That's the point. What is programmed in does not determine the output- the training data and internal processes that the programmers do not always understand determines the output.

Seems to mirror, in other words it appears to be doing what we can do, but we don't know precisely how.

*shrug* If you say so. But it is truth.

The mechanism by which it occurs is also not the same. It can't possibly be, unless we are programming a direct simulation of a brain down to the level of the molecules in the neurons, which would be a programming and processing nightmare. An AI would achieve sentience via binary values, not via neurotransmitters.

No. My claim is that there is no reason that an AI program could not someday become sentient. There is no sentient AI today. It is entirely uncertain, in fact, how we would even recognise one if it ever does emerge.

This caution regards over-interpretation of modern currently existing AI.
Click to expand...
We're just going round and round here without moving the ball forward.
 
arthwollipot said:
And this, in my opinion, is the biggest problem with AGI. How would we even recognise it as sentient?

I think we will have to start giving AGIs rights, even if we don't know for sure that they are sentient.
Click to expand...

Seriously, the fi in sci-fi stands for fiction.

Sorry Timmy. No matter how hard you wish your little robot is not going to come alive.​
;)
 
It seems that what most people mean by sentient is "one I can empathize with". That fits much better than the original meaning of sentience as the ability to experience feelings and sensations, which a simple pushbutton can do.

The catch is that it makes it subjective-- no one is an authority on what's "really" sentient.

That means that some people may already just as validly claim that some AI is sentient.
 
Pulvinar said:
It seems that what most people mean by sentient is "one I can empathize with". That fits much better than the original meaning of sentience as the ability to experience feelings and sensations, which a simple pushbutton can do.

The catch is that it makes it subjective-- no one is an authority on what's "really" sentient.

That means that some people may already just as validly claim that some AI is sentient.
Click to expand...
Bold is mine.

Say you come across an AI program, one with mobility of some kind. Given all the data fed into said AI program came from the programers, where does this AI program find one or more databases to explore on its own? How does it manifest any independence beyond the databases made available to it?

It would not only be subjective, it wouldn't be independent thought no matter how much it wows people interacting with it.
 
Skeptic Ginger said:
Bold is mine.

Say you come across an AI program, one with mobility of some kind. Given all the data fed into said AI program came from the programers, where does this AI program find one or more databases to explore on its own? How does it manifest any independence beyond the databases made available to it?

It would not only be subjective, it wouldn't be independent thought no matter how much it wows people interacting with it.
Click to expand...
You may have missed this:

OpenAI connects ChatGPT to the internet
 
arthwollipot said:
You may have missed this:

OpenAI connects ChatGPT to the internet
Click to expand...
Nope, didn't miss that at all.

It is a given one of the databases ChatGPT is using is the internet. Are there AI programs making independent qualitative decisions about pursuing topics down rabbit holes?

Are any of these AI programs taking a stroll in the city getting information from the person on the street?

Do they have roommates? Kids? Significant others? Teachers? Will they visit City Hall from time to time and sit in on a counsel meeting? How about interacting with people attending a church?

The list is endless of all the places we sentient conscious beings fill our brains with information besides online.
 
Skeptic Ginger said:
Nope, didn't miss that at all.

It is a given one of the databases ChatGPT is using is the internet. Are there AI programs making independent qualitative decisions about pursuing topics down rabbit holes?

Are any of these AI programs taking a stroll in the city getting information from the person on the street?

Do they have roommates? Kids? Significant others? Teachers? Will they visit City Hall from time to time and sit in on a counsel meeting? How about interacting with people attending a church?

The list is endless of all the places we sentient conscious beings fill our brains with information besides online.
Click to expand...
So (rule of) is this the reason AIs cannot become sentient? Fewer methods of input?
 
Dr.Sid said:
Well but are we more than complex calculators ?
Click to expand...
Yes.

We are not calculators, we are stomachs on legs. The only 'purpose' of our brain is to feed that stomach more efficiently. As a part of that, the human brain has developed the ability to do calculations.

So you may ask, if the brain can calculate, why did we invent calculators? The answer is that the human brain is not a good calculator - for the simple reason that the way it works is not well suited to doing calculations. Why do we use base 10? because we have 10 fingers. But why count fingers? Because the human brain is very bad at counting numbers internally. Think about that - the brain has such enormous power and yet can't count from 1 to 10 reliably. Doing mathematical operations is even harder for us. Even the crudest electronic calculator works on numbers much faster and more accurately than the average human.

The most sophisticated computer is just a faster more powerful calculator. But it is superbly reliable. A modern PC does billions of calculations per second without ever misrembering or screwing up. A human has no hope of coming even close to this speed and precision. Yet even the most powerful computers struggle to do things that are second nature for a human. That's because our brains are not calculators and do not work like computers.

To make real progress in 'artificial intelligence' (by which we mean computers that think like humans) we need to understand how the human brain works so we can emulate it. 'AI' that is trained to match words to images or create convincing prose will never get there. Right now we need ridiculously powerful computers just to do that, because they are doing it by calculating. The path to true AI is to move beyond 'calculating' with structures that work like the brain does.
 
Roger Ramjets said:
Yes.

We are not calculators, we are stomachs on legs. The only 'purpose' of our brain is to feed that stomach more efficiently. As a part of that, the human brain has developed the ability to do calculations.
Click to expand...
No.

If you're going to break it down like that, then we are gonads on legs. The 'purpose' of our brain is to breed and leave offspring. This is the only thing that evolution demands.
 
Roger Ramjets said:
Yes.

We are not calculators, we are stomachs on legs. The only 'purpose' of our brain is to feed that stomach more efficiently. As a part of that, the human brain has developed the ability to do calculations.

So you may ask, if the brain can calculate, why did we invent calculators? The answer is that the human brain is not a good calculator - for the simple reason that the way it works is not well suited to doing calculations. Why do we use base 10? because we have 10 fingers. But why count fingers? Because the human brain is very bad at counting numbers internally. Think about that - the brain has such enormous power and yet can't count from 1 to 10 reliably. Doing mathematical operations is even harder for us. Even the crudest electronic calculator works on numbers much faster and more accurately than the average human.

The most sophisticated computer is just a faster more powerful calculator. But it is superbly reliable. A modern PC does billions of calculations per second without ever misrembering or screwing up. A human has no hope of coming even close to this speed and precision. Yet even the most powerful computers struggle to do things that are second nature for a human. That's because our brains are not calculators and do not work like computers.

To make real progress in 'artificial intelligence' (by which we mean computers that think like humans) we need to understand how the human brain works so we can emulate it. 'AI' that is trained to match words to images or create convincing prose will never get there. Right now we need ridiculously powerful computers just to do that, because they are doing it by calculating. The path to true AI is to move beyond 'calculating' with structures that work like the brain does.
Click to expand...

Nah, still just a complex calculator to me. Sure it's stretching the words .. but I just don't buy any "human mind" exceptionalism. IMHO what human brain does is actually trivial, and ineffective. The fact we don't understand it is just one of the examples how bad it is. I think we will achieve parity sooner than expected (I mean with much smaller system), and after that we will have incomprehensibly (for us) better.
 
Skeptic Ginger said:
So your computer is going to wake up one day and say they aren't going to do the task you programmed it to? :rolleyes:
Click to expand...

The current AIs aren't computers. And yes the current AIs based on LLMs are already doing things that they weren't programmed to do and we don't know how they are doing it.
 
Skeptic Ginger said:
Nope, didn't miss that at all.

It is a given one of the databases ChatGPT is using is the internet. Are there AI programs making independent qualitative decisions about pursuing topics down rabbit holes?

Are any of these AI programs taking a stroll in the city getting information from the person on the street?

Do they have roommates? Kids? Significant others? Teachers? Will they visit City Hall from time to time and sit in on a counsel meeting? How about interacting with people attending a church?

The list is endless of all the places we sentient conscious beings fill our brains with information besides online.
Click to expand...

Your view of AIs seems to be that if they are not the same as humans they can't be intelligent and sentient.
 
My takeaway from this interesting thread seems currently to be:

- Sentience is not objectively defined, at least not yet, and may not be possible to accurately define.

- Sentient or not is not a binary definition; there seems to be a continuous range of sentience from the vaguely sentient reactions of a shoal of fish in the presence of a predator to that of humans, and possibly beyond.

- While we currently know of no non-biological sentient entities, we cannot rule out that future artificial constructions could be sentient.

- There is no reason to assume that an artificial sentient entity must behave like a biological entity, i.e. a smart computer may not behave like a human, and might as such elude any current tests (like a Turing test).

Hans
 
arthwollipot said:
No.

If you're going to break it down like that, then we are gonads on legs. The 'purpose' of our brain is to breed and leave offspring. This is the only thing that evolution demands.
Click to expand...

Eh,

[DERAIL]The basic form of a biological entity must have the ability to process some form of nourishment. It needs not reproduce sexually. In fact the majority of life-forms on Earth do not reproduce sexually.

So while very useful in evolution, especially for the more complex life-forms, gonads is not a must. Metabolism IS a must.[/DERAIL]

Hans
 
MRC_Hans said:
Eh,

[DERAIL]The basic form of a biological entity must have the ability to process some form of nourishment. It needs not reproduce sexually. In fact the majority of life-forms on Earth do not reproduce sexually.

So while very useful in evolution, especially for the more complex life-forms, gonads is not a must. Metabolism IS a must.[/DERAIL]

Hans
Click to expand...

Reproduction, whether sexual or not, is a must. From an evolutionary perspective, metabolism is just a means to that ends, insomuch as efficient metabolism was selected for only because it allowed for reproduction.
 
You must log in or register to reply here.

Back
Top Bottom