Consciousness: what it is, where it comes from, do machines can have it and why do we care?

[Somebody613]

I would say we want the most "simple" animal that shows evidence for an abstract idea. If rats playing computer games with their brains and ChatGPT "hallucinating" is not abstract enough I do not know what what sort of evidence we are going to get though.

Anything that isn't directly rooted in the ability to interact with the physical world around the test subject.
The rats example is still physical, though - and I never claimed that animals CAN'T have stuff like photographic memory leading to environment mapping.
That's still NOT what I'm trying (and failing) to describe as "the point of the quantum leap in intelligence".

 

[Snowygerry]

I would say we want the most "simple" animal that shows evidence for an abstract idea. If rats playing computer games with their brains and ChatGPT "hallucinating" is not abstract enough I do not know what what sort of evidence we are going to get though.

Burial gifts, offerings, non-descriptive art idk. ?

 

[Akka]

So what we're looking for is the first evidence of an abstract idea ?

If we go at the root of the, well, concept, an abstract idea is about whatever a mind conceives that is not a direct perception/reaction to the physical world. Even extremely simple and basic thoughts can fit the requirements.
Admitedly, there is a grey area when it comes to such basic thought, where it's unclear if it's comprehensively manipulating a concept, or just hardwired instinct (an insect perceiving and obstacle and going around it for example).

But anything above that (any evidence of understanding rather than just perceiving) is a proof of mental abstraction. So any animal able to learn is evidence of abstract idea - you can't "learn" something physical, by its very essence learning is an entirely mental process that deals with ideas.

 

[Somebody613]

But anything above that (any evidence of understanding rather than just perceiving) is a proof of mental abstraction. So any animal able to learn is evidence of abstract idea - you can't "learn" something physical, by its very essence learning is an entirely mental process that deals with ideas.

You are contradicting yourself. When a bee learns (and even teaches) any new "maps" it saw some useful flowers - is it "perceiving" or "understanding" the positions of cars and pencils?
Because I can assure you that cars aren't a "naturally found" object, so a "wild" bee would know exactly zilch about them, until and unless it LEARNED what it is.
So how is a bee different from a rat, then?
Both LEARN new stuff -and then even TEACH it to others.
See, where you get stuck with "perception"?
Unless you WANT to say that bees are as smart as mammals - I mean, I say they ARE (because neither is ACTUALLY "smart", but just "programmed" into behavioral patterns, lol).

 

[Samson]

Anything that isn't directly rooted in the ability to interact with the physical world around the test subject.
The rats example is still physical, though - and I never claimed that animals CAN'T have stuff like photographic memory leading to environment mapping.

From the writeup:

The rats were then placed on a treadmill ball within a 360-degree immersive virtual reality (VR) arena.​

​

In a subsequent experiment, the team gave the rats a “Jedi task” in which the animals themselves were stationary but had to direct an object on the screen to a particular goal within the VR environment using only their brain activity. Once again, the rats were able to do so.​

And the abstract:

Rats first formed a hippocampal map of a virtual environment. Then, in brain‐machine interface mode, they demonstrated the ability to activate representations from this map corresponding to specific remote locations, which then brought either them or an object to spatial goals. The rats could sustain a hippocampal representation of a remote location for tens of seconds, reminiscent of human imagination or mental time travel.​

What exactly is "physical" here? What would we have to remove to make it abstract?

You are contradicting yourself. When a bee learns (and even teaches) any new "maps" it saw some useful flowers - is it "perceiving" or "understanding" the positions of cars and pencils?
Because I can assure you that cars aren't a "naturally found" object, so a "wild" bee would know exactly zilch about them, until and unless it LEARNED what it is.
So how is a bee different from a rat, then?
Both LEARN new stuff -and then even TEACH it to others.
See, where you get stuck with "perception"?
Unless you WANT to say that bees are as smart as mammals - I mean, I say they ARE (because neither is ACTUALLY "smart", but just "programmed" into behavioral patterns, lol).

If bees are conscious then it probably is at the hive level. Who knows on what level they perceive, understand and learn. I see no objective way to rule it out it in such a different organism.

 

[Akka]

You are contradicting yourself.

No.

When a bee learns (and even teaches) any new "maps" it saw some useful flowers - is it "perceiving" or "understanding" the positions of cars and pencils?

The very concept of "map" is abstract. It requires building a mental construct about different positions of things you aren't seeing at the moment, ergo it's manipulating abstract ideas (ideas about the things you're locating, ideas about their relative positions).
That a bee is not able to understand what non-flower are is irrelevant to the fact that it can build a mental construct about the position of flowers it detected.
I already pointed this to you, that you ignore it is your fault, not mine, you just keep proving you're not really grasping the concepts you're talking at length about.

 

[Somebody613]

The very concept of "map" is abstract.

Then suggest me a better word for something like "beauty" or "happiness", because those are NOT physical or even mental.
Those are properties that are merely perceived when attached to physical objects, but this attachment in no way defines the property itself.
Not too dissimilar (in a certain sense) from how "electricity" is NOT "moving electrons", but rather something we "attach" to the explanation built upon "moving electrons".
It's still a property that can't be "touched" without thirst attaching it to ANOTHER entity.
Conversely, "objects in memory" are still precisely the same physical objects (as concepts), merely being "perceived" via a different medium (not eyes, but "inner eyes" of imagination).
On the other hand, "chess-checkers" is an example of breaking this pattern, because it takes the concept of a "chess piece" and then transforms it into the concept of a "checkers piece".
Neither a "chess piece" nor a "checkers piece" is BY ITSELF a physical object (because you can "play" using bottle caps, lol, which are NOT "game pieces"), so it's already step one.
But it takes a much more complex step two to "replace" the already non-physical concept of a "checkers piece" with the equally non-physical concept of a "chess piece".
Now, at SUCH a level - I'm sufficiently convinced that we are finally talking about "abstract NON-physical entities" and intellectual capabilities.
But, fine, help me name it better, if you think that "abstract" starts at a much more "physical" step.

 

[Samson]

Burial gifts, offerings, non-descriptive art idk. ?

Not really what you asked, but this point brought me to this paper that has an interesting model of awareness of death. Is death abstract? Can we put a "consciousness" zone on this:


Integrated model of life–death awareness. The first levels are governed by perceptual categorisation, whilst the second and third levels are governed by associative concept learning and high-order reasoning (analogical/inductive/causal reasoning), respectively. Species possessing all these cognitive processes are in a likely position to acquire an emergent conceptual awareness of death.

 

[Akka]

Then suggest me a better word for something like "beauty" or "happiness", because those are NOT physical or even mental.
Those are properties that are merely perceived when attached to physical objects, but this attachment in no way defines the property itself.
Not too dissimilar (in a certain sense) from how "electricity" is NOT "moving electrons", but rather something we "attach" to the explanation built upon "moving electrons".
It's still a property that can't be "touched" without thirst attaching it to ANOTHER entity.
Conversely, "objects in memory" are still precisely the same physical objects (as concepts), merely being "perceived" via a different medium (not eyes, but "inner eyes" of imagination).
On the other hand, "chess-checkers" is an example of breaking this pattern, because it takes the concept of a "chess piece" and then transforms it into the concept of a "checkers piece".
Neither a "chess piece" nor a "checkers piece" is BY ITSELF a physical object (because you can "play" using bottle caps, lol, which are NOT "game pieces"), so it's already step one.
But it takes a much more complex step two to "replace" the already non-physical concept of a "checkers piece" with the equally non-physical concept of a "chess piece".
Now, at SUCH a level - I'm sufficiently convinced that we are finally talking about "abstract NON-physical entities" and intellectual capabilities.
But, fine, help me name it better, if you think that "abstract" starts at a much more "physical" step.

There is no better word because the differences you're talking about are difference of complexity, not nature. "beauty" is an abstract concept, just like "left" or "right", and there is no fundamental gap between them that would require a different word.
(also I'm afraid to tell you, your examples are absolutely dreadful at illustrating your points ; "happiness" is an emotional state, so in fact pretty directly tied to concrete perception, and using chess pieces to play checkers doesn't require any subsequent steps in abstraction, in fact it's even less abstract than using chess pieces to play chess)

Most animals with a brain are able to manipulate abstract concepts - that's needed once you go anywhere beyond simple hardwired reaction to stimuli. Being able to evaluate danger is already about patterns recognition after all.
From there, it's just "more complex", not "fundamentally different" if we go by the intelligence scale (there is probably vastly more fundamental difference between a crocodile and a cow than between a cow and a human in fact).
You can of course set up tresholds to differenciate some sort of significant steps in intellectual ability (like the ability to plan for the future or the ability to understand death), but they would simply be about "complex enough" rather than "not working the same".

And about your desire to create a fundamental difference between "humans" and "animals", good luck. The most intelligent animals are already pretty close, if not overlapping slightly, with the least intelligent humans, so you'll have a tough job building a wall here.

 

[Somebody613]

and using chess pieces to play checkers doesn't require any subsequent steps in abstraction, in fact it's even less abstract than using chess pieces to play chess

Here's where I see why we think differently.
Using bottle caps as checkers, yes, that'd be the same as checkers itself.
But point (and task for AI) was to take what is DEFINED as "chess" and then use it what is DEFINED as "checkers".
The point being that the player "detaches" the "chess-ness" from the pieces and "attaches" the "checker-ness" to them.
It's clearly more complex than merely "attaching" the "checker-ness" to, say, bottle caps (which are not "another game's pieces").
Basically, it's "two steps" -vs- "one step", or so I see it.
But that's just my opinion, so whatever.

 

[Akka]

Here's where I see why we think differently.
Using bottle caps as checkers, yes, that'd be the same as checkers itself.
But point (and task for AI) was to take what is DEFINED as "chess" and then use it what is DEFINED as "checkers".

Using chess pieces to play checkers is not different than using rocks to play checkers. You take some objects and define them as "checker pieces". It's not like chess pieces have some inherent property that makes them more complex to use as checker piece than said rocks.
What would require more abstraction would be if using chess pieces would alter the checker rules in some way specific to chess pieces. That's not the case.

The point being that the player "detaches" the "chess-ness" from the pieces and "attaches" the "checker-ness" to them.
It's clearly more complex than merely "attaching" the "checker-ness" to, say, bottle caps (which are not "another game's pieces").
Basically, it's "two steps" -vs- "one step", or so I see it.
But that's just my opinion, so whatever.

No, you're just jumping through weird and unneeded hoops to find complexity where there is none. I could just as well say that I need to "detach" the "bottlecapness" from bottlecap and then attach the "checkerness". See, two steps, just like for chess pieces.

 

[Moriarte]

Cool thread though. No one has good enough idea of what is discussed.

But point (and task for AI) was to take what is DEFINED as "chess" and then use it what is DEFINED as "checkers".

Something like this?

Chapaev: Checkers Battle play online

Russians have made up a very interesting checkers game, named for Vasily Chapaev. The rules are very easy and remind the shuffle game.

web.archive.org

 

[Somebody613]

Using chess pieces to play checkers is not different than using rocks to play checkers. You take some objects and define them as "checker pieces". It's not like chess pieces have some inherent property that makes them more complex to use as checker piece than said rocks.
What would require more abstraction would be if using chess pieces would alter the checker rules in some way specific to chess pieces. That's not the case.

No, you're just jumping through weird and unneeded hoops to find complexity where there is none. I could just as well say that I need to "detach" the "bottlecapness" from bottlecap and then attach the "checkerness". See, two steps, just like for chess pieces.

I was talking about a test for a chess-AI in the first place, and for it, "chess-ness" is actually a thing.
For a human, of course, it's like you said - but that's not the task I was referring to.

Alternatively, the way *I* play chess-checkers (I literally based this on my own experience, lol) - I use King/Queen/Rooks as "(primary) crowned checkers", so I kinda DO apply "residual chess-ness" to the "now-checker" pieces.

 

[Somebody613]

Something like this?

Chapaev: Checkers Battle play online

Russians have made up a very interesting checkers game, named for Vasily Chapaev. The rules are very easy and remind the shuffle game.

web.archive.org

Not quite. Chapaev's pieces are still checkers (or rather, "bottle caps" technically) - and I highly doubt you can actually PLAY (well, effectively) Chapaev with chess pieces, loool.

 

[Birdjaguar]

You are contradicting yourself. When a bee learns (and even teaches) any new "maps" it saw some useful flowers - is it "perceiving" or "understanding" the positions of cars and pencils?
Because I can assure you that cars aren't a "naturally found" object, so a "wild" bee would know exactly zilch about them, until and unless it LEARNED what it is.
So how is a bee different from a rat, then?
Both LEARN new stuff -and then even TEACH it to others.
See, where you get stuck with "perception"?
Unless you WANT to say that bees are as smart as mammals - I mean, I say they ARE (because neither is ACTUALLY "smart", but just "programmed" into behavioral patterns, lol).

Sailboats are not a natural object found in oceans and yet beginning around 2020 orca pods in the north Atlantic have "invented" a way they can sink them. We don't know why they do so, only that they can and do sink them without trying to injure the humans on board. This takes planning, implementation and communication skills before an attack and then coordinated action around the boat. The attacks all focus on the rudder area.

 

[Somebody613]

Sailboats are not a natural object found in oceans and yet beginning around 2020 orca pods in the north Atlantic have "invented" a way they can sink them. We don't know why they do so, only that they can and do sink them without trying to injure the humans on board. This takes planning, implementation and communication skills before an attack and then coordinated action around the boat. The attacks all focus on the rudder area.

You got my point in reverse. I'm saying that you can see this type of LEARNED and TAUGHT behavior already in BEES. So how come you think mammals are "smarter", lol?

 

[Birdjaguar]

@Somebody613 About beauty and happiness. You imply that such notions exist independent of the physical world in some ethereal place. Our brains are very hard wired into responding to signals from the physical world through our senses. I would contend that our brains experience happiness and beauty from physical stimulus. Our brains have response mechanism that create those feelings. Having those distinct experiences has enabled people to generalize the concepts of beauty and happiness into the abstract. Certainly dogs and cats experience happiness and seek it out. Whether or not they have abstracted the notion is beyond our skill to discover as yet. So I say both beauty and happiness as ideas have their roots in the physical experiences and we have just made up that there is an abstract model.

 

[Birdjaguar]

You got my point in reverse. I'm saying that you can see this type of LEARNED and TAUGHT behavior already in BEES. So how come you think mammals are "smarter", lol?

Orcas were not taught by anyone nor did they learn it. They invented it on their own and unlike bees, which you say do not incorporate unnatural objects in their mapping, orcas have included man made boats. Boat sinking has never been part of orca behavior before. They invented it all on their own. In addition, so far boat sinking is only found among orcas in the North Atlantic. We'll have to see if it spreads to other pods over time.

EDIT: If it spreads, then it is likely that whales are teaching other whales how to do it and why. That is how culture works. Someone has a new idea; they put it into practice to refine it and then then teach others why it is important and how to do it.

 

[Somebody613]

Orcas were not taught by anyone nor did they learn it. They invented it on their own and unlike bees, which you say do not incorporate unnatural objects in their mapping, orcas have included man made boats. Boat sinking has never been part of orca behavior before. They invented it all on their own. In addition, so far boat sinking is only found among orcas in the North Atlantic. We'll have to see if it spreads to other pods over time.

EDIT: If it spreads, then it is likely that whales are teaching other whales how to do it and why. That is how culture works. Someone has a new idea; they put it into practice to refine it and then then teach others why it is important and how to do it.

Dude, you clearly read me backwards.
Bees DO incorporate the mapping of objects that they SHOULDN'T understand as OBJECTS in the first place: cars, glass windows, appliances.
Which means that they interact with reality around them JUST LIKE rats do.
So, are you now gonna say that bees are AS SMART AS rats?
Because, well, you can say so easily - but then it utterly deconstructs the whole "evolution mentality arrogance" of "being the smartest mammal".
If a bee can do it - it certainly doesn't need to "evolve" a zillion more steps into a mammal, which then makes the entire "arrogance" of "higher development brain", well, false, lol.

Orcas already cooperate between each other (unless you meant that they do it in "single player") - that's the very definition of TEACHING someone, lol.
Though I don't find this example of any special use, given how my current question is about BEES to begin with.

 

[Somebody613]

About beauty and happiness.

I may have used a wrong example in happiness, since it's an emotion that comes from within the user.
Beauty, though, is not the emotion itself, but rather a "judgment" of an external "info packet" - basically, we find something to be "beautiful" REGARDLESS of how "useful" it is to us.
A beautiful painting of a sunset has exactly zero "physical benefits" for the observer - yet it causes a specific reaction in them that can't be measured beyond the binary yes/no.
Okay, there may be degrees of "beauty", but they are also "reaction-gauged" - you have exactly zero ways to explain why picture A is "more beautiful" than picture B.
No, not why you LIKE one more than other - but rather WHY you like one in the first place.
There MAY be some "physical" reason behind it in SOME cases - but that's definitely not ALWAYS there, and probably is a minority anyways.
Heck, WHY is a flower "beautiful", while a charcoal is "not" (assuming you feel that way)?
I bet you won't be able to EXPLAIN it beyond "I just FEEL this way".
Well, congrats, you are now using a NON-PHYSICAL concept, because you can't describe (aka gauge, aka define) it in a PHYSICAL way, beyond merely stating your feelings about it.
The end, lol.