A Workable Definition Of Intelligence

[Max Faust]

With respects to the concept "artificial intelligence" (AI) there are some fuzzy edges, to say the least. What is intelligence? Is it defined by cultural parameters? Being "smart" is very likely to mean something completely different for a machine (AI) than for a man. I would say that it follows from this that a hypothetical alien life form which has a stucture and strategy for self maintenance that is vastly different from ours is likely to have a system of abstract cognition which is also vastly different from ours. Perhaps all meaningful communication is left a practical impossibility?

I'd say that it's possible to visualise a type of intelligence which isn't about manipulating your surroundings ("nature") at all, but rather about accentuating the efficiency of adaptation. This will make the idea that you need to have as well "culture" as advanced technology in order to qualify as "intelligent" obsolete.

In order to reach out for a common denominator: Is is fair to say that "intelligence" is about how you process information?

 

[guerom00]

I though that intelligence was the ability to adapt to new situations… No ?

 

[Max Faust]

Maybe something which includes both complexity and adaptability?

 

[apeiron]

This question has only just been asked in this thread...
https://www.physicsforums.com/showthread.php?t=382800

A way to naturalise or generalise the definition of intelligence lies in the idea of anticipation. That is, using past experience to predict (and thus hope to control) future events.

Computers, of course, are quite terrible at learning and anticipating. They are the dumbest things in creation really. Though there have been reasonable attempts to construct anticipatory neural networks (Grossberg ART nets in particular).

Anticipation is another way of saying "complexity and adaptability" perhaps. There is also the approach known as complex adaptive systems (CAS) which has similarities - such as hierarchical structure.

But a key point about anticipatory approaches is that it reverses the usual concept of processing. Instead of input being turned into output, the system generates the output first (it forward models, it creates a state of sensorimotor prediction). Then it responds to input according to the already prevailing state. So what was correctly predicted can be ignored.

It is the errors of prediction - the surprises - or the sought for matches (which are also surprises in that you were expecting the light to turn green, but you didn't know exactly where or when) which become the focus of attention, the focus of further processing. Learning results from novelty and so the system becomes better at predicting the next time round.

I'm getting off the track, but I just wanted to point out that intelligence is about how we (not computers) process information. And we do it the other way round, via anticipation or forward modelling.

This then leads to your query about possible alien super-intelligence that may be some pure form of cognition and not all messy with culture (and perhaps you are also thinking bodily needs and unnecessary emotions). This is the Dr Spock model of IQ.

Again, computer technology as we know it is immensely dumb. The best we can say for it, as with all machines, is that it serves to amplify human capabilities - including these days, especially with the internet, human cultural and social evolution.

Advanced technology is not more intelligent than advanced culture, merely a possible medium for its enlarged expression.

Another dimension to the discussion would be that intelligence always has a purpose. It is defined by what it can achieve (and goals are what it anticipates achieving).

So again, this idea of a cold, disemboddied, disinterested super-intelligence is a popular one in science fiction, but quite unrealistic from a theoretical biology point of view. Systems have to have purposes to exist (or rather, persist). Otherwise why would they even bother to be.

Any aliens would have to be shaped by the same general constraints as us. The same "laws of biology" would apply everywhere. And so we could feel assured that aliens would use anticipatory "processing", that they would be dissipative structures, that they would encode purposes, etc. They would be nothing like a computational "intelligence" in other words.

 

[MaitreyaB]

First off, I take no firm stand but lean toward the undefinability of the word 'Intelligence'

The way the human brain works is to recognize symbols and associate them*. The faster and more proper it does this, the more commonly it would be referred to as 'intelligent' behavior.

*For example: The smell of garlic is a symbol your mind receives in conjunction with later having a full stomach. Result: the smell of garlic leads your brain to think food is the next step. Think Pavlov's dog.

 

[zomgwtf]

First off, I take no firm stand but lean toward the undefinability of the word 'Intelligence'

No offense but from the very few posts I've seen you make you strike me as the type of person who believes philosophy is a game of semantics.

 

[DaveC426913]

First off, I take no firm stand but lean toward the undefinability of the word 'Intelligence'

The way the human brain works is to recognize symbols and associate them*. The faster and more proper it does this, the more commonly it would be referred to as 'intelligent' behavior.

*For example: The smell of garlic is a symbol your mind receives in conjunction with later having a full stomach. Result: the smell of garlic leads your brain to think food is the next step. Think Pavlov's dog.

Did you not just equate intelligence with the conditioned response of Pavlov's dog? That's kind of exactly not intelligence, isn't it?

 

[DaveC426913]

My definition of intelligence involves the ability to adapt existing principles to new situations.

 

[zomgwtf]

My definition of intelligence involves the ability to adapt existing principles to new situations.

Indeed not only that but to anticipate the changes. aperions post seems quite good.

 

[kote]

No offense but from the very few posts I've seen you make you strike me as the type of person who believes philosophy is a game of semantics.

Wittgenstein made that claim pretty strongly and influentially. Philosophy of language is a big issue, and a lot of contemporary analytical philosophy has been focused on language and the fact that it's imperfect and necessarily imprecise. Lewis (and that camp) didn't claim that language was basic, but it's certainly very important to pretty much all of his philosophy.

There are a lot of ways to look at the issue, but for the most part debates about what intelligence is or what life is are entirely semantic. Because of the imprecision in how we learn such terms, and our willingness to redefine words as we learn new things, we run into problems particularly at the fuzzy edges. Still - it's just semantics unless your definitions depend on other concepts. If you define living things as requiring a soul, for example, then you also need to figure out what a soul is and whether or not we have them etc.

 

[DaveC426913]

If you define living things as requiring a soul, for example, then you also need to figure out what a soul is and whether or not we have them etc.

https://www.physicsforums.com/showthread.php?t=391762

 

[lisab]

With respects to the concept "artificial intelligence" (AI) there are some fuzzy edges, to say the least. What is intelligence? Is it defined by cultural parameters? Being "smart" is very likely to mean something completely different for a machine (AI) than for a man. I would say that it follows from this that a hypothetical alien life form which has a stucture and strategy for self maintenance that is vastly different from ours is likely to have a system of abstract cognition which is also vastly different from ours. Perhaps all meaningful communication is left a practical impossibility?

I'd say that it's possible to visualise a type of intelligence which isn't about manipulating your surroundings ("nature") at all, but rather about accentuating the efficiency of adaptation. This will make the idea that you need to have as well "culture" as advanced technology in order to qualify as "intelligent" obsolete.

In order to reach out for a common denominator: Is is fair to say that "intelligence" is about how you process information?

I like the simplicity of the Turing test for AI. That is: if an examiner is chatting online and he can't tell if he's chatting with a person or an AI, then that computer has passed the Turing test.

Now, if you happen to be chatting with a computer that *aced* the Turing test, http://xkcd.com/329/" could happen.

 

[zomgwtf]

Wittgenstein made that claim pretty strongly and influentially. Philosophy of language is a big issue, and a lot of contemporary analytical philosophy has been focused on language and the fact that it's imperfect and necessarily imprecise. Lewis (and that camp) didn't claim that language was basic, but it's certainly very important to pretty much all of his philosophy.

There are a lot of ways to look at the issue, but for the most part debates about what intelligence is or what life is are entirely semantic. Because of the imprecision in how we learn such terms, and our willingness to redefine words as we learn new things, we run into problems particularly at the fuzzy edges. Still - it's just semantics unless your definitions depend on other concepts. If you define living things as requiring a soul, for example, then you also need to figure out what a soul is and whether or not we have them etc.

This is surely true, I don't disagree at all. However this isn't the same as saying that philosophy is a semantics game. At least when I had posted that it wasn't what I was intending it to look like.

 

[DaveC426913]

Now, if you happen to be chatting with a computer that *aced* the Turing test, http://xkcd.com/329/" could happen.

You know, before I even tried the link, my thought was "This sounds like an XKCD reference..."

 

[GreatEscapist]

Props to aperions post.
I'm not very philosophical-politically correct, but I would define intelligence also as the ability to adapt, comprehend, and overcome changes.
I would also say that is the ability to reason, find patterns, and solve problems. Which I guess could be adapting to your surroundings. To me, (please no one eat me for saying this) people who excel in math, science, and language are more intelligent than those that do not.

 

[Kajahtava]

Props to aperions post.
I'm not very philosophical-politically correct, but I would define intelligence also as the ability to adapt, comprehend, and overcome changes.
I would also say that is the ability to reason, find patterns, and solve problems. Which I guess could be adapting to your surroundings. To me, (please no one eat me for saying this) people who excel in math, science, and language are more intelligent than those that do not.

It implies that people drastically get less intelligent and that being moral aequals stupidity and that all children exceed their parents intelligence.

Which is what I believe myself by the way, but just pointing it out.

 

[GreatEscapist]

It implies that people drastically get less intelligent and that being moral aequals stupidity and that all children exceed their parents intelligence.

Which is what I believe myself by the way, but just pointing it out.

Moral doesn't equal stupidity. That helps you cope and comprehend what you do. It depends on how moral you are, and if it has grounds.
And otherwise- yes, that is what I'm saying. And I think it's typically true. But there are always exceptions to the rule.
That's the point of kids- to have the potential to be more intelligent than the parent. The how society evolves and gets smarter.

 

[binbots]

I just asked the same question about intelligence. "defining intelligence". Its seems intelligence is not something you can measure because everyone will think of intelligence as something different. Some see it as being good at math, some at biology etc. Some might see it as how well you can reproduce and how well your offspring do. Some poeple see it as how much money you can get. Some see it as how happy one self is.
But even if you all decide on one of these subjects do define intelligence then it still is undefind. If we all agree that people who know pyhsics are the most intelligent then the problem is who in physics is the most intelligent. People who study pyhsics will use a idea or a formula created by someone else. Is this smart or is it just having a good memory. Even if one comes up with a original idea there is a good chance that not all will agree upon it. There for it is not absolute intelligence. These forums are great prove of this. Not many here agree on anything there for no one is more intelligent than any other.
Even when people agree on something it is just a matter of time before it is disproven or a better idea comes along. So in the end I believe that there is no such thing as intelligence, only opinion and beleive.

 

[Kajahtava]

Moral doesn't equal stupidity. That helps you cope and comprehend what you do. It depends on how moral you are, and if it has grounds.
And otherwise- yes, that is what I'm saying. And I think it's typically true. But there are always exceptions to the rule.
That's the point of kids- to have the potential to be more intelligent than the parent. The how society evolves and gets smarter.

You said intelligence was the ability to accept change. That for a fact diminishes with age.

A baby will accept about every change there is, an adult with be resistant towards.

And any moral notion is a refusal to accept change, the only reason we believe some things are 'bad' is because we've been told they are that many times, often coupled with punishment when we didn't accept it, that we've come to believe such absurd nonsense that some things can be 'bad' or 'good', and the cycle goes on, and on, and on.

 

[DaveC426913]

You said intelligence was the ability to accept change. That for a fact diminishes with age.

Not any change!
Advantageous changes. Accepting any change is not a sign of intelligence.

 

[GreatEscapist]

And any moral notion is a refusal to accept change, the only reason we believe some things are 'bad' is because we've been told they are that many times, often coupled with punishment when we didn't accept it, that we've come to believe such absurd nonsense that some things can be 'bad' or 'good', and the cycle goes on, and on, and on.

Be careful when you use absolute words. Any moral notion is such a strong statement, and not right at all. Having moral notions doesn't always resist change. It sure makes it HARD sometimes, but not impossible to change. Sometimes, when used correctly, it helps you change and grow.

Not any change!
Advantageous changes. Accepting any change is not a sign of intelligence.

You should put that in a high school, especially that last sentence.

 

[apeiron]

Not any change!
Advantageous changes. Accepting any change is not a sign of intelligence.

And how do you know what to accept unless you are anticipating the consequences?

But again, this tacit input-output model is putting the cart before the horse.

My earlier point was to stress that the way to detect change is to expect something, then find something else happened. Intelligence can be considered as the thrashing about mentally to find some new approach that will more correctly predict the world in the future (problem-solving, really). Or it can be considered as all the accumulated habits of thought and perception which do an effective job of anticipating reality, minimising errors of prediction and other kinds of unpredicted surprises.

(What you do as an action in a particular moment vs what you develop as a capacity over sufficient time)

Children are indeed in the immature learning phase so far as this weight of accumulated anticipatory knowledge is concerned. Their responses may be more coarse-grained or otherwise clumsy and exploratory.

Grown-ups have adapted to their worlds and developed their various physical and mental competencies to the full. Further learning becomes fine-grained at best. We tend to talk about the transition from clever to wise.

All this is modeled in the neural network literature. I've given the references many times.

To remind, the OP was about alien intelligences that might be like computers and have no concept of culture. It is really more of a science question than a philosophical one. And there is really no shortage of science to apply here.

 

[apeiron]

If you want the same point about the inadequacy of input-output models of cognition (the ones physicists and computer scientists find "natural"!) repeated in someone else's words, then here is a good review article.

http://frontiersin.org/neuroscience/humanneuroscience/paper/10.3389/fnhum.2010.00025/html/

Although suggestions regarding the importance of predictive mechanisms originate from very early phases of both psychology and neurosciences, until recently they have not been strongly advocated in the majority of frameworks of cognitive and neural processing. Specifically, typical approaches in delineating cognitive processes postulate a rather serial process, starting with sensory, continuing with executive and “higher cognitive” functions and ending in overt behavior. Such thinking stems from the original behaviorist conceptualizations which emphasized the linear progression from sensory stimulation to overt behavior, a view which was also present in early information-processing cognitivist theories. And, even though the extreme behaviorist stimulus–response view of human behavior is today rarely or almost never advocated, different cognitive processes are today still dominantly studied in isolation. In addition, although rarely explicitly postulated, it is often assumed that a given process of interest starts with the output of some earlier, lower-level process and terminates once it provides input to the next processing stage.