What is Intelligence?

Discussions on general biology and biological evolution, genetics, zoology, ecology, botany, etc.

What is Intelligence?

Postby hyksos on March 20th, 2018, 6:33 am 

What is Intelligence?

An objective measurement of intelligence would apply to not only differences in people, but also be neatly extrapolated to non-human primates, elephants, and perhaps other mammals like dolphins.

If we had an objective measurement of intelligence -- meaning the efficacy of the brain organ -- we would have quantified it already. If we would have quantified it already, we would have built it already into an artificially intelligent agent.

It is true that the military wants that thing -- and wants it badly. The state of AI and robotics stands in as the strongest evidence that us humans are having a terrible time defining what "intelligence" means to begin with. Or it may be the case that we know exactly what it is, but we don't know how it works in brains.
robottrips.jpg


https://thumbs.gfycat.com/MemorableConcernedGypsymoth-size_restricted.gif


We do not possess a Theory of Intelligence like anything approaching the efficacy of a brain organ or its alleged "horsepower". If we had such a theory, the DoD would have put the engineers in a bunker in Los Alamos, and we would have robot tanks already. We don't.

https://en.wikipedia.org/wiki/List_of_animals_by_number_of_neurons
User avatar
hyksos
Active Member
 
Posts: 1318
Joined: 28 Nov 2014


Re: What is Intelligence?

Postby BadgerJelly on March 20th, 2018, 11:48 am 

Without computer aid and the ability to communicate in complex ways would Stephen Hawkings' intelligence been fulfilled.

If I am born with a terrible voice yet want to sing professionally can I still sing and reach the top? Yes, but I won't get to the top because I am unable to hits notes others hit with ease.

Do scientists measure intelligence? Yes.

What is it they are measuring? Look up the g-factor, it is not a new idea nor one that has gone away. It is also applied to animals.

Do the number of neurons dictate intelligence? Obviously I think we can say that generally speaking more neurons indicate more potential. Intelligence is also inhibited by access to information and some creatures have not been able to manipulate their environments to the extent that humans can.

Generally I would say empathy is massively important (theory of mind is an absolute must for intelligence.)

It doesn't matter how many bricks you have it matters how they are put together. An isolated infant human will quickly become nothing more than a pile of rubble even if provided for. If you take the "social" capacity out of the human you no longer have a human.

I don't really see a need to measure human intelligence to other animals. It is hard enough to distinguish cultural differences within human groups and we're all very similar at the genetic level.

I believe the g-factor has been observed in other species such as rats and mice?

You've defined "intelligence" in the OP and then asked what intelligence is. If you say it is efficiency then it is efficiency. I am guessing you're referring to processing speed here? I don't think that is really a measurement of intelligence, but it is certainly part of the package because something has to be processed in order to express intelligent behavior.

I have come across the same problem with defining "language." In the animal kingdom of linguistics it is perfectly acceptable to refer to bees as having a language, yet in common parse we don't really mean anything like human language - incidently neuroscience has opened up this area and the last I read humans exhibit language features that are a combination of properties exhibited in other animals, like birds. We just happen to have a particular admixture of language proclivity, where other animals have maybe one or two we have the "full package" (so to speak!)
User avatar
BadgerJelly
Resident Member
 
Posts: 5302
Joined: 14 Mar 2012


Re: What is Intelligence?

Postby wolfhnd on March 20th, 2018, 12:30 pm 

"The ability to acquire and apply knowledge and skills." Oxford Dictionary

Defining intelligence is not the problem observing it is.
User avatar
wolfhnd
Resident Member
 
Posts: 4628
Joined: 21 Jun 2005
Blog: View Blog (3)


Re: What is Intelligence?

Postby BadgerJelly on March 20th, 2018, 12:42 pm 

Here is a reasonable overview:

https://www.youtube.com/watch?v=oaJ01Ex7DLw

On a nuts and bolts basis, meaning on the biology of intelligence, there is not a lot to go on.

What the biological basis for the "g factor" is is up for debate. All the evidence points to the "g factor" being something more than a mere myth.
User avatar
BadgerJelly
Resident Member
 
Posts: 5302
Joined: 14 Mar 2012


Re: What is Intelligence?

Postby Serpent on March 20th, 2018, 1:34 pm 

hyksos » March 20th, 2018, 5:33 am wrote:What is Intelligence?

An objective measurement of intelligence would apply to not only differences in people, but also be neatly extrapolated to non-human primates, elephants, and perhaps other mammals like dolphins.

It does. The intelligence of birds and mammals is tested in quite an array of interesting ways. Most of these are non-verbal, so that the subject does not first have to acquire human language. https://phys.org/news/2017-05-scientists-probe-dolphin-intelligence-interactive.html and something similar for rodents http://learnmem.cshlp.org/content/15/7/516.full I've already cited this one for birds http://www.audubon.org/news/bird-iq-tests-8-ways-researchers-test-bird-intelligence

If we had an objective measurement of intelligence -- meaning the efficacy of the brain organ -- we would have quantified it already. If we would have quantified it already, we would have built it already into an artificially intelligent agent.

If we had space travel capability, we would already be living on Mars?
No, but some quite clever people are working on it, but every process takes time. The behaviorists have only just died out. Except for Konrad Lorenz, almost nobody considered other species capable of thought until the last couple of decades.

Obviously, all these tests are biased to human comprehension. The environments in which the tests take place are rigidly controlled and human-contrived - utterly foreign to the test subject. Nevertheless, many of the test subjects perform beyond the experimenters highest expectation - I suspect a good deal better than most humans would in the animal's environment.
We can never test other species quantitatively for facilities humans don't possess. But we now have the technical ability to follow them in their natural habitat and record behaviour when animals confront their real-life problems. A good deal of data have already been collected. Some of it has been put to human-serving use; much of it it still just unexploited science.

We do not possess a Theory of Intelligence like anything approaching the efficacy of a brain organ or its alleged "horsepower". If we had such a theory, the DoD would have put the engineers in a bunker in Los Alamos, and we would have robot tanks already. We don't.

We don't, but that doesn't mean they don't.
A theory is the beginning of a long process that summits as military deployment, then devolves through industrial application and household gizmos to children's toys.

For a short definition, I can live with "facility for learning". That can be compared to other individuals and other species.
https://boingboing.net/2015/01/15/are-you-smarter-than-a-rat.html
I'm not a fan of mazes for testing intelligence, since it leans heavily on a sense of direction, which I don't see linked to intelligence. I'm not even sure how much memory should be weighted-for. I much prefer the practical problem solving experiments.
It's still a young science, but it's growing.
Serpent
Resident Member
 
Posts: 3015
Joined: 24 Dec 2011


Re: What is Intelligence?

Postby hyksos on March 21st, 2018, 8:47 am 

Except for Konrad Lorenz, almost nobody considered other species capable of thought until the last couple of decades.

This is completely true. Good luck trying to explain it to BadgerJelly , though.

We don't, but that doesn't mean they don't.

This is a nonsensical conspiracy theory so I'm under no obligation to respond to it. I hope that you will not adopt the position that since I cannot prove the non-existence of secret AI robot tanks, therefore the military could already have an armada of them. (that would be a fallacy on your part).

The bleeding edge of AI is still basically Google R&D and Demis Hassabis and Andrew Ng and the rest of their et als. In 2018 AI is at the level of doing 'okay' on most 8bit ATARI games, and doing superhumanly well on a handful of them.

It is illuminating and ironic -- Absolutely no one in this thread has been able to articulate any reason as to why the robots shown above are so horribly bad at doing basic things in 2018.
User avatar
hyksos
Active Member
 
Posts: 1318
Joined: 28 Nov 2014


Re: What is Intelligence?

Postby BadgerJelly on March 21st, 2018, 10:37 am 

Hyksos -

No need to explain to me what I already know? I think you are stuck here. You've made a huge error in judgement. I was simply saying we're better off looking at comparisons between humans because we understand the human world - we're not ravens.

I think what I mentioned about language as an appropriate comparison. Mant people stick to hard and fast definitions of "language" whilst in the anima world linguists are more willing to broaden their use of the term to suit the capacities of different species.

What do I need convincing about?
User avatar
BadgerJelly
Resident Member
 
Posts: 5302
Joined: 14 Mar 2012


Re: What is Intelligence?

Postby BadgerJelly on March 21st, 2018, 10:47 am 

Hyksos -

It is illuminating and ironic -- Absolutely no one in this thread has been able to articulate any reason as to why the robots shown above are so horribly bad at doing basic things in 2018.


Perhaps someone would've tried if you'd asked the question? Just saying ...
User avatar
BadgerJelly
Resident Member
 
Posts: 5302
Joined: 14 Mar 2012


Re: What is Intelligence?

Postby hyksos on March 21st, 2018, 1:49 pm 

"Commonly accepted notions of X" are a terrible metric for gauging and explaining how brains work. That which is "commonly referred to as intelligence" is likely "smartness" and is therefore a strongly culturally-dependent notion that is dependent on the fads and fashion of the provincial society.

When rubber hits roads, and you have to actually publish about a how the mouse's cortex works... or when you really need to build these skills into a robotic agent. --- suddenly and violently these cultural "norms" have no use at all. There are not even a rough guideline about how to proceed. One gets a strong sense they are likely based on culture and have no basis in science.

Smarter and more talented people have already been over this treaded ground. They are called Eliminative Materialists.

https://plato.stanford.edu/entries/materialism-eliminative/

Now I'm not demanding that anyone become an immediate convert to Eli-Mat. But I would point out a very important aspect of one tenet of the tradition. Folk psychology.

Folk Psychology

This concept of "folk psychology". We all think we know how the brain works, and we have all wive's tales about brains work. And when you really push these things, they have little or nothing to do with the way the brain actually works. I sense in some people on this forum a dangerously enamored of pop psychology and its associated wive's tales. There is little appreciation (or knowledge of) how vastly ignorant science really is about how brains function.

There is a way we talk about the brain in a conversation over beer or coffee.. and then there is the way the brain actually functions. The gulf between these two things is wide and yawning.

I mean pop psych-ers start yammering about how "we know what intelligence is" and at the end of the day they are talking about a collection of outward behaviors. They have no idea how beholden their thinking really is to behaviorism. They speak of "complex problem solving ability" like radical behaviorists of the 1950s. They have no neuroscientific basis for what they are saying or what they are claiming... there is never any concessions to functional brain areas, for example.

They still discuss memory like it is a gigantic hard drive, or a repository like balls placed into a bucket. But memory does not act like this in brains at all. In fact the way memory functions in brains is excruciatingly complex and very little of it is really understood by science.

To give you an idea of how difficult the topic of Memory really is --- you have to first understand that no artificial intelligent agent in existence has Episodic Memory. Like zero of them do. THe entire discipline of AI lacks any kind of coherent or plausible algorithm to implement Episodic memory. They are some wild-haired nutty professors at the boundaries of academia who claimed to have solved it. But there is no consensus or set of established algorithms for it.

Furthermore, crows have no cortex. Yet they have exhibited complex tool-building abilities, seemingly concocted on the fly, that are simply astonishing to witness.

Because crows lack a cortex, their brains are more closely related to frogs and amphibians than they are related to mammals. Because of this, science has no theory at all about how crows reason abotu bending wires. The part of the brain that allows mammals to "complex problem solve" and bend a wire into a hook is literally physically absent in the crow's head.

Since science is not in possession of such a theory -- we can be resolutely certain that SOME GUY ON THIS FORUM has no theory either. Nor does Stephen Pinker.

Nobody at MIT, Stanford, or Cornell know how the crow is doing this.

A very large supercomputer project is being implemented in Europe, and involved international teams of neuroscientists. The project is called Blue Brain Project , and is being conducted in Geneva. One part of this research is to try to uncover the exact function of a part of the cortex called a cortical column. Science does not know what this networked unit is doing exactly in a brain. Some have suggested they are little "feature detectors" but that is an overweening conclusion drawn from computer science geeks who are enamored of machine learning. We know that evolution has tried to pack as many of these columns as it can fit into the human head. This is in fact the reason why our brains have folds in them.

When the smoke clears, we are going to have to have gained a better appreciation of the breathtaking ignorance that science has about brains and how they function. To keep defining the function of brains in mere outward behaviors has already been an approach that was heavily and widely investigated. That approach is called Behaviorism.

Despite our current ignorance, the last 18 years has seen an absolute explosion in neuroscience. A veritable golden age is at hand. This means, among other things, that we were even more ignorant of the brain in the 1970s, when IQ tests were first being hoisted and formulated as standardized tests for children.

To turn and run towards IQ tests as a safe ground for measuring the function of a brain organ is not to go forwards.. it is going backwards into a darker and more ignorant time. You are running backwards into a dark ages of neuroscience, when the brain was an absolute black box. Psychologists of that time had nothing to go on but outward behaviors that they supposed could be back-extrapolated to brain function. Today we see how woefully and terribly wrong they were. Outward behavior is NOT a good guide to brain function, at all. For in such a back-extrapolation you fall victim to folk psychology to try to explain what you are witnessing.

The utter failure to explain the crow behaviors in terms of brain regions, shows how sourly wrong this type of behaviorist thinking can go.

The first step in making progress in science is to first honestly coral and outline one's ignorance. But people on this forum can'[t make that first crucial step. They prefer to slap their hands over their ears and keep spouting contextless Stephen Pinker quotes -- like a person yelling to cover up the nakedness of their own ignorance.

It's high time to stop and join the rest of civilized society of the 21st century.
User avatar
hyksos
Active Member
 
Posts: 1318
Joined: 28 Nov 2014
Dave_C liked this post


Re: What is Intelligence?

Postby BadgerJelly on March 21st, 2018, 3:01 pm 

Have you had a recent break-up or something? You are sounding more and more unhinged by the day?

You are creating a false picture of what is being said. It is not my fault you wish to paint me a certain shade in order to push your point home.

Ignorance is my "god" btw. I love it and try to expand it. If you take offensive at some simply words about how "intelligence" is defined in psychology go ahead. The quote you are referring to is from the APA uttered by Pinker. You stated no one in psychology would say what the APA said; thus I proved your point to be little more than bombast (and most of the above is the same.)

May I ask if any neuroscientist refers to "intelligence"? If so what are they referring to?

The title of your OP is asking "What is Intelligence?", do you mean to say there is no "intelligence" and if not what is it?

Don't worry I don't expect much more than evasion and more name calling. If it helps you get over whatever has recently happened in your life I'm happy to be your punch bag :)

I am also curious what you'd make of wacky suggestions from neuropsychologists talking about studies in cognitive neuroscience where they use IQ test questions to study brain functioning. And if you'd be up in arms about them referring to something called G? (Of course such Folk Psychology must and use of terms like "episodic memory" and such Magical terminology must be utter trash right?)

ANYWAY, spit it out. What is intelligence? Does it exist? What does the up to the mark cognitive neuroscientist of the day proclaim about the issue? My ignorance in this area is vast, I make no denial of that in the slightest. I've merely read bits and bobs piece meal from a few text books and watched a few online lectures on the subject.

Teach me, show me where to look and what to read. I am willing enough to learn. Any links references or recommendations fo reading will be taken on board - already made note of the book you've mentioned elsewhere.
User avatar
BadgerJelly
Resident Member
 
Posts: 5302
Joined: 14 Mar 2012


Re: What is Intelligence?

Postby Serpent on March 21st, 2018, 3:29 pm 

hyksos » March 21st, 2018, 7:47 am wrote:
[We don't, but that doesn't mean they don't]
This is a nonsensical conspiracy theory so I'm under no obligation to respond to it.

You mean the military have no secret projects and devices?
I didn't expect a response; I merely pointed out, in a semi-joking manner, that what "we" have or know is not necessarily all that exists.

It is illuminating and ironic -- Absolutely no one in this thread has been able to articulate any reason as to why the robots shown above are so horribly bad at doing basic things in 2018.

Possibly because robots were not under consideration by anyone else? Personally, I have no experience of robotics and nothing useful to contribute on the subject.
Serpent
Resident Member
 
Posts: 3015
Joined: 24 Dec 2011


Re: What is Intelligence?

Postby hyksos on March 22nd, 2018, 4:49 am 

Serpent » March 21st, 2018, 11:29 pm wrote:Possibly because robots were not under consideration by anyone else? Personally, I have no experience of robotics and nothing useful to contribute on the subject.

I'm going to interpret this to mean that the problems existing in the discipline of robotics are to be considered separate and tangential to the hurdles within artificial intelligence. That's certainly an interesting perspective. Personally, I would like to see that perspective fleshed out in a little more detail.
User avatar
hyksos
Active Member
 
Posts: 1318
Joined: 28 Nov 2014


Re: What is Intelligence?

Postby hyksos on March 22nd, 2018, 5:18 am 

The ultimate goal of all this yammering about intelligence is to uncover an objective and principled Theory of Intelligence. This theory would be abstract enough to fit a large umbrella over almost all organisms found in nature that are endowed with a brain. It would be unifying enough that it would relate all the various competing AI approaches as "limiting cases" of the larger theory.

This Theory of Intelligence would apply equally well to robotics as it would to any of the other pure data-driven domains.

It would be best here to lay out on the table all the various existing approaches and compare and contrast them in terms of their assumptions and approximations. I was thinking that a colorful diagram would be a way to begin the process. (Maybe that is not needed yet) This following list is kind of off the top of my head, so I don't know all the researcher's name tip-of-tongue. Should serve well as an overview. others can fill in the gaps.

(>)
Intelligence should be abandoned, and replaced by consciousness. So what was traditionally a hierarchy of intelligence is actually a degree of how integrated an organism's conscious awareness is. Proponents : Giulio Tononi, Cristof Koch , Jeffrey Krichmar , Gerald Edelman

(>)
"The brain is just a prediction machine." (cr. Jeff Hawkins, and some other recent robotocists whose name I'm too lazy to look up.) This doctrine claims that what the brain does is produce predictions of the future, and then compare the difference (or "error") between the prediction and what actually happens. An agent or organism is more "intelligent" if the error of its predictions is very low.

(>)
"Intelligence is the generality of learning -- or the ability to transfer learning in one domain to a different domain." This is the AGI Doctrine. Its adherents include Demis Hassabis, Yann LeCun , Shane Legg, Marcus Hutter , Ben Goertzel, Hugo DeGaris

(>)
"Intelligence is about making decisions and taking actions under uncertainty." This is the Bayesian Doctrine. Adherents include : Peter Norvig, Stuart Russel, Andrew Ng, Sebastian Thrun, Eliezer Yudkowsky, Derek Muller.
Last edited by hyksos on March 22nd, 2018, 5:27 am, edited 1 time in total.
User avatar
hyksos
Active Member
 
Posts: 1318
Joined: 28 Nov 2014


Re: What is Intelligence?

Postby hyksos on March 22nd, 2018, 5:22 am 

8888888
User avatar
hyksos
Active Member
 
Posts: 1318
Joined: 28 Nov 2014


Re: What is Intelligence?

Postby BadgerJelly on March 22nd, 2018, 5:41 am 

Would it be possible to measure intelligence without testing the subject. I don't see how.

If we can say that memory is certainly important to intelligence then we could perhaps test and assess memory, short term and long term. Then we could measure the speed someone could order pieces of information too.

Simply pick some qualities that you believe are part of intelligence and then see what does and doesn't correlate. Break down the cognitive abilities into as many categories as you can (if you can.)

Even if we can do this it doesn't take into account a multitude of other influences upon how a person lives their lives.

So, first of all define "intelligence" and then go about figuring out how to assess it. Then once you've got something see how far it spreads into other human cognitive activities - and then there is the issue of unconscious processing (but that would be another thing to consider in terms of other species.)
User avatar
BadgerJelly
Resident Member
 
Posts: 5302
Joined: 14 Mar 2012


Re: What is Intelligence?

Postby hyksos on March 22nd, 2018, 6:53 am 

hen once you've got something see how far it spreads into other human cognitive activities - and then there is the issue of unconscious processing (but that would be another thing to consider in terms of other species.)

We have seen that in primates and even some great apes, that some species can recognize themselves in a mirror, while others cannot. These results challenge the doctrine of "Brain is a prediction machine". What is lacking on one species and the next is not merely a higher and lower degree of prediction from past experience. Its closer to : does not have key concept C.

Problematically, the mirror test suggest that intelligence is related to "higher abstract concepts" being present or not. I'm open to this type of theory myself, but this theory is very much out-of-fashion in the prevailing climate of cog sci.
User avatar
hyksos
Active Member
 
Posts: 1318
Joined: 28 Nov 2014


Re: What is Intelligence?

Postby Asparagus on March 22nd, 2018, 8:45 am 

The mirror test (self recognition) is taken to signify intelligence, isn't it?

Octopi pass it, which for some reason blows my mind.
Asparagus
Member
 
Posts: 259
Joined: 16 Dec 2017
Blog: View Blog (2)


Re: What is Intelligence?

Postby Braininvat on March 22nd, 2018, 1:17 pm 

I think the key word here is "umbrella." Intelligence is an umbrella term (reminiscent of Wittgenstein's family of resemblances) that, from my perspective, boils down to a range of abilities. There are many sorts of ability and therefore anything that could be called "general intelligence" would involve the capacity to quickly shift from using one ability to using another. Or use several together to solve problems, discover new facts about reality, predict what will happen, etc. Abilities might include such things as were mentioned in above posts, like predicting, deciding, transferring learning from one domain to another, decoding abstract symbol systems, perceiving the mental states of other beings, modeling various outcomes of an action before taking it, revising models of reality based on new information as it is presented, making causal connections that accurately reflect the true state of affairs, constructing coherent narratives that share easily with others, and so on.

In another thread, I suggested that aptitude tests (which measure such specific abilities) are of more value than trying to find some sort of mean of general intelligence and then arranging tidy little points around it. Whenever you try to quantify that umbrella, you inevitably create a bias and a tendency to give more weight to certain kinds of flashy intellectual pursuits. If AIs ever develop something that we could call general intelligence, and ace their Turing Tests, it won't be because someone engineered anything as nebulous as general intelligence, it will be because all those aptitudes and abilities I described above have converged and synergized in a way that successfully makes its way in the world.
User avatar
Braininvat
Forum Administrator
 
Posts: 6564
Joined: 21 Jan 2014
Location: Black Hills


Re: What is Intelligence?

Postby Serpent on March 22nd, 2018, 2:05 pm 

hyksos » March 22nd, 2018, 3:49 am wrote:[Possibly because robots were not under consideration by anyone else? Personally, I have no experience of robotics and nothing useful to contribute on the subject.]
I'm going to interpret this to mean that the problems existing in the discipline of robotics are to be considered separate and tangential to the hurdles within artificial intelligence.

Why would you interpert it that way? I meant only what I said; nothing bigger. I don't know what hurdles AI faces. I don't know whether a General Theory of Intelligence is one of those hurdles, or whether it's even a good idea for AI to succeed.
My gut feeling is, AI will evolve by alternate fertile and static periods, like everything else, and that nobody working on it today know what will happen next year or in a decade.

I tend not to trust any of the current methods of measuring intelligence (that I know of) because they're all anthropocentric. Like the mirror test. While humans are endlessly fascinated by their own image - to the extent of fashioning deities and nature-spirits in the same semblance - no other species in its natural habitat is interested in looking at itself.
Many other animals, and even more so, birds and fish, make visual displays in order to elicit some particular response from other member of their species, but that's not calculated behaviour; it neither requires nor indicates intelligence.
When dogs try to put something over on their masters by looking abashed or sorry, by feigning a limp or pretending indifference, that is calculated behaviour and does show intelligence: the ability to observe the reaction of a completely different species, relate it to circumstances, relate it to one's own actions, predict what affect that reaction will have on oneself, conceive a desired outcome, plan an artificial action, devise a mode of conveying the specially designed message that other species will understand. That's an astounding series of abstract mental operations. But it's impossible to measure on a standard scale.
All the tests we give them are artificial - unrelated to the normal life of that animal. And yet so many members of so many species are able to adapt to the human media of communication and perform human-devised tricks.
I think that indicates more than intelligence; it indicates a high degree of flexibility... soft-wiring, if I may hijack a bad metaphor.

Similarly, when some academic in Boston devises an IQ test for a goatherd in Bali, he may think he's adjusted for culture, but it's the Balinese who has to come far out of his comfort-zone, even just to attempt the ridiculous test, never mind make sense of the problems. The academic has no clue what sort of problems are natural to the subject, and yet he's the one judging the other's intelligence.

That's the only kind of data we have, and I consider it inadequate to draw any conclusions. However, with improvements in candid video transmission, we can discover much more about the natural behaviour, learning and problem-solving capability of other species.
Still, I'd just as soon not subject the Balinese or Cuicas to that kind of scrutiny.
Serpent
Resident Member
 
Posts: 3015
Joined: 24 Dec 2011


Re: What is Intelligence?

Postby BadgerJelly on March 23rd, 2018, 12:50 am 

You may find this interesting in relation to exploration and intelligence (look 47:00, rise of dopergenic reward):

https://www.youtube.com/watch?v=LOY3QH_jOtE&index=15&list=PLD7E21BF91F3F9683

When something changes we're more interested in it. Novelty is something we're primed to investigate. This is why I am addicted to studying :)

How does this play into what we call "intelligence" and memory access?
User avatar
BadgerJelly
Resident Member
 
Posts: 5302
Joined: 14 Mar 2012


Re: What is Intelligence?

Postby dandelion on March 25th, 2018, 6:59 pm 

I think problems could involve possibly anthropocentric bias in ToM-like conceptions.
dandelion
Member
 
Posts: 355
Joined: 02 May 2014


Re: What is Intelligence?

Postby Katrin on March 28th, 2018, 4:22 am 

From the proteins, archaea-and cyanobacterias, the first kick started life to
thinking humans:
A smart development looking for improvements and problem-solutions for solve the existence through targeted multiplication
with something or someone that is conductive and necessary for improvement
User avatar
Katrin
Forum Neophyte
 
Posts: 9
Joined: 14 Jul 2016
Location: NorthGermany


Re: What is Intelligence?

Postby BadgerJelly on March 28th, 2018, 4:56 am 

dandelion » March 26th, 2018, 6:59 am wrote:I think problems could involve possibly anthropocentric bias in ToM-like conceptions.


What does "ToM-like" mean?
User avatar
BadgerJelly
Resident Member
 
Posts: 5302
Joined: 14 Mar 2012


Re: What is Intelligence?

Postby Braininvat on March 28th, 2018, 9:43 am 

Theory of Mind
User avatar
Braininvat
Forum Administrator
 
Posts: 6564
Joined: 21 Jan 2014
Location: Black Hills


Re: What is Intelligence?

Postby dandelion on March 28th, 2018, 1:51 pm 

Yes, Theory of Mind, thanks, Biv, and reading this page more thoroughly, I think broader and well explained similar sentiments were already posted by Serpent.
dandelion
Member
 
Posts: 355
Joined: 02 May 2014


Re: What is Intelligence?

Postby BadgerJelly on March 28th, 2018, 8:21 pm 

The problem is defining what you wish to measure. That is all. We're not about to give a worm an IQ test, and we are not concerned with a worms intelligence.

I think we have non-starter in what hyksos proposes. There appears to be a concern with rating the intelligence of other species by finding a neural mechanism. The question then becomes whether or not we know if someone can swim by looking at their brain. The better test would be to throw them into water and see for ourselves.

We know more about human intelligence and consciousness so why conflate the measurement into other species? We don't measure distances by the length of an ant.
User avatar
BadgerJelly
Resident Member
 
Posts: 5302
Joined: 14 Mar 2012


Re: What is Intelligence?

Postby Serpent on March 28th, 2018, 11:07 pm 

BadgerJelly » March 28th, 2018, 7:21 pm wrote:The problem is defining what you wish to measure. That is all.

Not quite all. If you define liquid and want to know how much of it you have, you still need to figure out the appropriate measuring device. A ruler won't do. A graduated cylinder does measure its volume, but you still need a scale to measure its weight, and a hydrometer to measure its density. And even with three different devices, you still won't know every quantitative aspect of the fluid.

We're not about to give a worm an IQ test, and we are not concerned with a worms intelligence.

Speak for yourself. I find it an interesting subject.

I think we have non-starter in what hyksos proposes. There appears to be a concern with rating the intelligence of other species by finding a neural mechanism.

Seeing as how thought is a product of the brain, and the quality and quantity of thought is a reflection of the quality and quantity of brain producing it, this seems reasonable.

The question then becomes whether or not we know if someone can swim by looking at their brain. The better test would be to throw them into water and see for ourselves.

That seems unnecessarily cruel. Still, the presence of fins might provide a clue as to the likelihood of a creature's ability to swim. Its buoyancy might provide another. Appraisal of the available equipment is not an unreasonable starting-point for assessing the probability of an aptitude.

We know more about human intelligence and consciousness so why conflate the measurement into other species?

We know something about human intelligence. That doesn't give all the information we could possible want regarding all intelligence.

We don't measure distances by the length of an ant.

We measure distance by the length of some arbitrarily chosen object that we agree to use a standard. Why is a dead king's lower extremity more appropriate to use than an ant?
Serpent
Resident Member
 
Posts: 3015
Joined: 24 Dec 2011


Re: What is Intelligence?

Postby BadgerJelly on March 28th, 2018, 11:43 pm 

You've still expressed elsewhere what information or understand you have of this subject so I can only regard your posts as mere speculation and unfounded opinion.

And again, you still need to define what you're measuring. The ability to solve problems is measured by giving people problems to solve and comparing them to each other. By all means look at the neural structuring too, but this doesn't tell us if someone is intelligent or not (as far as I know.) And given that neuroscientists use IQ tests and questions to "look for" intelligence I would say that is something worthy of note wouldn't you?

For example, they don't ask someone to sing a song in order to look for intelligence any more than they ask them to ride a bike while they take readings.

That seems unnecessarily cruel. Still, the presence of fins might provide a clue as to the likelihood of a creature's ability to swim. Its buoyancy might provide another. Appraisal of the available equipment is not an unreasonable starting-point for assessing the probability of an aptitude.


I am not sure if you're being deliberately facetious here. You fail to address the "aptitude". Decide on what you're measuring before you even look at the brain. Isn't that obvious?
User avatar
BadgerJelly
Resident Member
 
Posts: 5302
Joined: 14 Mar 2012


Re: What is Intelligence?

Postby Serpent on March 29th, 2018, 1:50 am 

BadgerJelly » March 28th, 2018, 10:43 pm wrote:You've still expressed elsewhere what information or understand you have of this subject so I can only regard your posts as mere speculation and unfounded opinion.

OK. It's only natural that, if you don't know my foundations and can't see rationality in my statements, you should assume I have no foundation, but simply pull opinions out of thin air.

And again, you still need to define what you're measuring.

I don't need to define a word if I'm content with the one in my dictionary. I'm content with Webster's
"a : the ability to learn or understand or to deal with new or trying situations. b : the ability to apply knowledge to manipulate one's environment or to think abstractly "; have no objection to Braininvat's
"the capacity to quickly shift from using one ability to using another. Or use several together to solve problems, discover new facts about reality, predict what will happen, etc. " and accept Wolfhnd's succinct definition as the ability to learn and apply knowledge.
I don't think definition is an obstacle. I don't think philosophy is an obstacle.
I don't even think attempts at quantifying is an obstacle, so long as we bear in mind the limitation of our testing and measuring methods.

The ability to solve problems is measured by giving people problems to solve and comparing them to each other.

Within clearly understood parameters. That clear understanding is not invariably present.

By all means look at the neural structuring too, but this doesn't tell us if someone is intelligent or not (as far as I know.)

It doesn't tell us how intelligent someone is compared to other someones. It does suggest whether intelligence is likely. The relative sophistication of the hardware gives a clue to the probability of higher or lower intelligence. These are clues - not conclusions.

And given that neuroscientists use IQ tests and questions to "look for" intelligence I would say that is something worthy of note wouldn't you?

Answered umpteen times, yes, of course. IQ tests are useful, within the limits of the application.

[ Appraisal of the available equipment is not an unreasonable starting-point for assessing the probability of an aptitude. ]
I am not sure if you're being deliberately facetious here. You fail to address the "aptitude".

How is it facetious to link fins and buoyancy with the ability to swim? That was your example. I didn't think it needed more addressing

Decide on what you're measuring before you even look at the brain. Isn't that obvious?

No, it's not obvious. In most of life, we see things long before we decide what we're looking for. The world is there before we know we understand anything about it or have any words for it. We learn as we go along. We make connections as we observe phenomena.
We don't determine patterns until after we've discerned patterns. We don't define concepts until after we've become aware of them. In language, we usually have a long-standing usage, shared and generally understood among the speakers of that language, long before a standardized official definition is articulated. I accepts all the definitions of intelligence in common usage. I accept all the means of measurement. With reservations.
Serpent
Resident Member
 
Posts: 3015
Joined: 24 Dec 2011


Re: What is Intelligence?

Postby BadgerJelly on March 29th, 2018, 2:51 am 

Serpent -

My mistake. I meant to type "You have not expressed" what information you're working with before when I've asked.

So, yeah. It makes perfect sense to assume you're dropping unfounded opinions here when you simply do not drop anything to back up what you're saying other than a dictionary reference.

Even then, the question still remains exactly how are you going to measure:
" the ability to learn or understand or to deal with new or trying situations. b : the ability to apply knowledge to manipulate one's environment or to think abstractly "; have no objection to Braininvat's "the capacity to quickly shift from using one ability to using another. Or use several together to solve problems, discover new facts about reality, predict what will happen, etc. " and accept Wolfhnd's succinct definition as the ability to learn and apply knowledge.


I have mentioned "crystalline" and "fluid" intelligence before. Have you any idea how such things could be measured or way in which they are actually applied to tests?

There are case studies where people have received brain lesions and lost most of their fluid intelligence yet remain perfectly able to solve complex questions like those they've come across before (because they are working from their "crystalline" intelligence.)

In terms of memory I think I have pointed out the difference in "explicit" and "implicit" memory ... then again I may have posted that video on another link?

We can say, to some degree, that the number of neurons matters from species to species, but the issue is more about how they are connected and the bodily functions (which would include social interactions and an ability to manipulate the environment.) As I have contrasted intelligence with language, we are well aware (or at least I am) that no exposure to language does not mean an inability to learn a language - this is what I would take away from the study posted elsewhere in these forums of the deaf man from Mexico who was first exposed to what we call "language" when he was ... 29 I believe? I made this contrast because "intelligence" is not something we pick up it is something exposed, like language.

The function of exploration of the environment is connected to comfort and stress (something I have mentioned several times.) Feeling in control effects how we perform. This is one area where I find IQ test circumspect. The degree of relation between neuroticism and intelligence has been noted in some studies.

I don't actually think that personality traits are separate from intelligence. That is, that personality traits are not cognitive functions. Necessarily I would say that some personality traits (to some degree) play into the "exposure" of intelligence because they are inextricable from the g-factor, just as abstract thought requires some form of symbolic representation, or an "as if" mode of thinking.

I am not inclined to look to bundle "intelligence" under the umbrella of "consciousness" in order to pretend it is an issue purely for the science of consciousness, any more than I am willing to suggest consciousness can exist without intelligence. Again, we can run around in circles trying to define these things or set out purposeful delineations in order to make some kind of scientific progress ... this has been the course of action in the cognitive sciences and the rise of neuroscience.

I get the feeling people are more concerned with the nut and bolts than with what they do together. There seems to me to be a disjoint between the actual day-to-day psychological understanding of humans and the neuroscience, when in fact they are reliant upon each other to further understanding. It is not a A versus B situation.

I am happy to go with "uncertainty" as being part of what intelligence deals with. If this is accepted then neuroticism plays a role, and the question becomes is neuroticism a condition of intelligence, or if some aspects of neuroticism should be put under the heading of "intelligence" in order to create a better measure?

The problem is there are layer upon layer of environmental factors so it appears the best we've got to date is evidence of a g-factor that shines through all intelligence tests. If we wish to look for an underlying neural mechanism for "intelligence" then we'll likely have to make multiple studies on various different tests and see what common neurons light up in the brain.

Then if you wish to ignore possible effects of reading someones brain you'll have to take into account the psychological effects that may bias such experimentation. We're left with approximations.

As for theory of mind, when it comes ot fluid intelligence I believe octopuses are more "intelligent" than humans on some tests. They require one viewing to replicate a task - although there is debate over whether they really "know" and questions over their capacity for crystalline intelligence I believe?

I hope we can agree on one thing. That is computers are not intelligent. I would say "consciousness" is a prerequisite for "intelligence." I would therefore say that "a capacity to learn" does not in and of itself represent "intelligence,' even though I would with equal vigor say that "a capacity to learn" is required for intelligence" - and here I openly admit to making a required delineation between the differences of "fluid" and "crystalline" intelligence as being made out of convenience (that being more a question of "consciousness" not intelligence.

In this sense we could say, at a stretch, that computers possess ONLY "crystalline" intelligence, and that my definition of intelligence requires BOTH. Therefore computers are not intelligent.

No matter what we do we are constantly EMOTIONALLY involved in the world. I cannot see any possibility for "intelligence" without th epotential for complex emotional interactions (be it with people, other animals or simply rocks and sticks.) That is irrefutable no matter how anyone chooses to dress it up - hence my inclination to look at personality traits in respect to intelligence.
User avatar
BadgerJelly
Resident Member
 
Posts: 5302
Joined: 14 Mar 2012


Next

Return to Biology

Who is online

Users browsing this forum: No registered users and 8 guests