AI that isn't intelligent.

[DrZoidberg]

Here's a really interesting talk. This is a cognition researcher who is also a game designer. She explains why it's a waste of time trying to create actors in games that are autonomic. It's much easier to fool us into thinking something is acting intelligently than actually making it act intelligently. And she goes through methods for doing so. This is very practical advice targeted at other game designers.

This is interesting for several reasons. How often do you project intelligence onto others or onto things in the world? Do you even know? Can you even know? She explains how stereotyping works and that we need to stereotype to function in the world. But it also means that everybody is always wrong. We all generalise too much in order to truly understand the world.

https://www.youtube.com/watch?v=1xWg54mdQos

The nice thing about this talk is that she can prove everything and she's got access to metrics to back it up. It's also games. So we can play these games and see for ourselves.

Good talk. Recommend it.

 

[Bronzeage]

Acting as if one has intelligence, instead of being intelligent, has always worked for humans.

 

[Speakpigeon]

It's much easier to fool us into thinking something is acting intelligently than actually making it act intelligently.

This reminds me of ants. Social insects have always fascinated me and I think all of mankind. Their performance really comes out of non-intelligent collective behaviours. Yet I think we feel it is nonetheless somehow intelligent. Now, if you compare a beehive or an ant colony to our brain there not much difference. So is our supposed intelligence not a little bit overrated?
EB

 

[J842P]

It's much easier to fool us into thinking something is acting intelligently than actually making it act intelligently.

This reminds me of ants. Social insects have always fascinated me and I think all of mankind. Their performance really comes out of non-intelligent collective behaviours. Yet I think we feel it is nonetheless somehow intelligent. Now, if you compare a beehive or an ant colony to our brain there not much difference. So is our supposed intelligence not a little bit overrated?
EB

Perhaps the *system* of ants is intelligent, rather than the individual ants themselves. Biological demarcations of objects are pretty arbitrary. Peter van Inwagen posits that the only objects are either elementary particles or living organisms. I think he is almost right - the only objects are particles.

 

[fromderinside]

... the only objects are particles fields.

Fixed it.

 

[J842P]

... the only objects are particles fields.

Fixed it.

A field is a mathematical abstraction, which represents something. But sure, whatever.

 

[DrZoidberg]

This reminds me of ants. Social insects have always fascinated me and I think all of mankind. Their performance really comes out of non-intelligent collective behaviours. Yet I think we feel it is nonetheless somehow intelligent. Now, if you compare a beehive or an ant colony to our brain there not much difference. So is our supposed intelligence not a little bit overrated?
EB

Perhaps the *system* of ants is intelligent, rather than the individual ants themselves. Biological demarcations of objects are pretty arbitrary. Peter van Inwagen posits that the only objects are either elementary particles or living organisms. I think he is almost right - the only objects are particles.

I agree. Ant colonies are "super organisms". Their intelligence is collective. Beehives are the same. We shouldn't be fooled by the fact that there's a physical separation between each ant. There's separation between the cells in our bodies, but they manage to coordinate just fine.

An argument can be made that the entire planet Earth is just one huge super organism. This is how I see it.

 

[bilby]

Perhaps the *system* of ants is intelligent, rather than the individual ants themselves. Biological demarcations of objects are pretty arbitrary. Peter van Inwagen posits that the only objects are either elementary particles or living organisms. I think he is almost right - the only objects are particles.

I agree. Ant colonies are "super organisms". Their intelligence is collective. Beehives are the same. We shouldn't be fooled by the fact that there's a physical separation between each ant. There's separation between the cells in our bodies, but they manage to coordinate just fine.

An argument can be made that the entire planet Earth is just one huge super organism. This is how I see it.

Eusociality can lead to colonies that are reasonably described as super organisms. Bees, ants, mole-rats, etc., maybe can be considered super organisms in that vein.

But the idea that the entire biosphere is a super organism is not so reasonable. Eusocial behaviour requires more than mere interaction between individuals; it requires close kinship that makes individual sacrifice evolutionarily beneficial. It's a nice idea, but it doesn't bear close examination.

It's very popular with the tree-hugging crowd, of course. But a super organism requires more than just "hey, everything's connected, man".

 

[DrZoidberg]

I agree. Ant colonies are "super organisms". Their intelligence is collective. Beehives are the same. We shouldn't be fooled by the fact that there's a physical separation between each ant. There's separation between the cells in our bodies, but they manage to coordinate just fine.

An argument can be made that the entire planet Earth is just one huge super organism. This is how I see it.

Eusociality can lead to colonies that are reasonably described as super organisms. Bees, ants, mole-rats, etc., maybe can be considered super organisms in that vein.

But the idea that the entire biosphere is a super organism is not so reasonable. Eusocial behaviour requires more than mere interaction between individuals; it requires close kinship that makes individual sacrifice evolutionarily beneficial. It's a nice idea, but it doesn't bear close examination.

It's very popular with the tree-hugging crowd, of course. But a super organism requires more than just "hey, everything's connected, man".

It comes down to definitions ultimately. But everything is connected. There tiny differences between us and almost everything else that's alive.

And I think stuff like Wikipedia and forums like this can be defined as a human hive mind.

 

[Speakpigeon]

... the only objects are particles fields.

Fixed it.

Good peasant common sense.
EB

 

[Speakpigeon]

This reminds me of ants. Social insects have always fascinated me and I think all of mankind. Their performance really comes out of non-intelligent collective behaviours. Yet I think we feel it is nonetheless somehow intelligent. Now, if you compare a beehive or an ant colony to our brain there not much difference. So is our supposed intelligence not a little bit overrated?
EB

Perhaps the *system* of ants is intelligent, rather than the individual ants themselves.

Yep, but so our intelligence is really environment-based. It works well within a given environment, or probably a class of environments, just as the ant colony's intelligence works very well only as long as you don't put it in an environment too different from their actual natural environment.

Biological demarcations of objects are pretty arbitrary. Peter van Inwagen posits that the only objects are either elementary particles or living organisms. I think he is almost right - the only objects are particles.

I disagree.
EB

 

[Speakpigeon]

I agree. Ant colonies are "super organisms". Their intelligence is collective. Beehives are the same. We shouldn't be fooled by the fact that there's a physical separation between each ant. There's separation between the cells in our bodies, but they manage to coordinate just fine.

An argument can be made that the entire planet Earth is just one huge super organism. This is how I see it.

Eusociality can lead to colonies that are reasonably described as super organisms. Bees, ants, mole-rats, etc., maybe can be considered super organisms in that vein.

But the idea that the entire biosphere is a super organism is not so reasonable. Eusocial behaviour requires more than mere interaction between individuals; it requires close kinship that makes individual sacrifice evolutionarily beneficial. It's a nice idea, but it doesn't bear close examination.

It's very popular with the tree-hugging crowd, of course. But a super organism requires more than just "hey, everything's connected, man".

Life taken as a whole must be really more intelligent than mankind if it has been able to adapt itself to an ever changing environment for hundreds of million of years. I think we all have doubts as to whether mankind will be able to go beyond a few more millenia, if that.
EB

 

[Juma]

Fixed it.

A field is a mathematical abstraction, which represents something.

So is particles.

 

[bilby]

Eusociality can lead to colonies that are reasonably described as super organisms. Bees, ants, mole-rats, etc., maybe can be considered super organisms in that vein.

But the idea that the entire biosphere is a super organism is not so reasonable. Eusocial behaviour requires more than mere interaction between individuals; it requires close kinship that makes individual sacrifice evolutionarily beneficial. It's a nice idea, but it doesn't bear close examination.

It's very popular with the tree-hugging crowd, of course. But a super organism requires more than just "hey, everything's connected, man".

Life taken as a whole must be really more intelligent than mankind if it has been able to adapt itself to an ever changing environment for hundreds of million of years. I think we all have doubts as to whether mankind will be able to go beyond a few more millenia, if that.
EB

To consider life as a whole 'intelligent' requires a highly idiosyncratic definition of intelligence. I don't think the concept applies. Life is adaptable, but that is a necessary, rather than a sufficient, condition for something to be considered intelligent. Evolution through natural selection is the epitome of an unintelligent adaptive mechanism - the bits that don't fit simply die.

As for mankind, reports of our imminent extinction have been greatly exaggerated. We might die off in large numbers, but as highly numerous generalists who already occupy a very wide range of environments, we will be hard to eradicate - even if we put a fair effort into trying to commit species suicide. Extinction of humanity (or even of all life on Earth) is a popular fear, and is trotted out fairly often in an attempt to persuade others to behave as we demand; but it is not actually very likely to happen, except as a result of unavoidable external events over which humans have no control, such as a giant meteor impact, or a Solar fluctuation.

If we managed to wipe out 99.999% of the world population, we would still have ~80,000 people with whom to start over - which is more than the hypothesised world population at the time of the Toba bottleneck of ca. 70,000 years ago. Genetic evidence suggests that genus Homo may have fallen to as few as 2,000 individuals prior to the beginning of the Palaeolithic. So you need to get population down below that number before you can assume that extinction is inevitable. Even a massive nuclear exchange would likely leave at least that many survivors in remote locations.

 

[Speakpigeon]

Life taken as a whole must be really more intelligent than mankind if it has been able to adapt itself to an ever changing environment for hundreds of million of years. I think we all have doubts as to whether mankind will be able to go beyond a few more millenia, if that.
EB

To consider life as a whole 'intelligent' requires a highly idiosyncratic definition of intelligence.

Hell, no! We've just worked very hard to generalise the concept of intelligence out of its idiosyncratic anthropocentred acceptation. Please pay attention.

I don't think the concept applies.

Then give us some more acceptable definition.

Life is adaptable, but that is a necessary, rather than a sufficient, condition for something to be considered intelligent.

you'd be more convincing if you could give us the other conditions necessary to something being intelligent.

Evolution through natural selection is the epitome of an unintelligent adaptive mechanism - the bits that don't fit simply die.

See our exchange about ants. Like the individual ants, the bits are not intelligent but the whole of life is--like the ant colony.

As for mankind, reports of our imminent extinction have been greatly exaggerated. We might die off in large numbers, but as highly numerous generalists who already occupy a very wide range of environments, we will be hard to eradicate - even if we put a fair effort into trying to commit species suicide. Extinction of humanity (or even of all life on Earth) is a popular fear, and is trotted out fairly often in an attempt to persuade others to behave as we demand; but it is not actually very likely to happen, except as a result of unavoidable external events over which humans have no control, such as a giant meteor impact, or a Solar fluctuation.

If we managed to wipe out 99.999% of the world population, we would still have ~80,000 people with whom to start over - which is more than the hypothesised world population at the time of the Toba bottleneck of ca. 70,000 years ago. Genetic evidence suggests that genus Homo may have fallen to as few as 2,000 individuals prior to the beginning of the Palaeolithic. So you need to get population down below that number before you can assume that extinction is inevitable. Even a massive nuclear exchange would likely leave at least that many survivors in remote locations.

I absolutely agree with the general idea. It's not plausible that anything we could do now could effectively exterminate humanity any time soon. Even Trump won't.

However, I didn't suggest that extinction was imminent. And your reasoning only really applies for relatively short period of times, say of the order of less than 100,000 years. Beyond that, we really don't know. And the fact that humanity already did survive near-extinction (in a very different environment) doesn't mean it will always do.

I also didn't suggest that extinction was "inevitable". I suggested that extinction was likely.

I suggested it was more plausible that extinction will occur within some hundreds of million of years than not.


If only out of boredom.
EB

 

[fromderinside]

Just a thought. Anything that is part of something is less than the sum of that something. Whatever the configuration for intelligence is finally agreed upon it must satisfy the above requirement.

 

[DBT]

The brain could be described as a colony of cells forming connections and groups in order to produce something greater than its individual parts (as with ants), an intelligent system.

 

[fromderinside]

back to greater than sum of parts again. Sheesh.

Something composed of parts is different than, not greater than, the sum of parts since when properly assembled that something is the sum of those parts.

Even more limiting parts when combined with other parts have many possible manifestations. Only of a few of those manifestations are ever exhibited when something is particularly combined into something meaning that the potential sum of parts is much more than a particular sum of parts.

More than sum of parts is merely a description of a thing being since the parts uncombined are not realizing all potential combinations. An overreach and an err when elevated to justification for the existence of something like consciousness as a fundamental force or particle since its dependence on being properly assembled from parts to be at all.

 

[DBT]

back to greater than sum of parts again. Sheesh.

Something composed of parts is different than, not greater than, the sum of parts since when properly assembled that something is the sum of those parts.

Even more limiting parts when combined with other parts have many possible manifestations. Only of a few of those manifestations are ever exhibited when something is particularly combined into something meaning that the potential sum of parts is much more than a particular sum of parts.

More than sum of parts is merely a description of a thing being since the parts uncombined are not realizing all potential combinations. An overreach and an err when elevated to justification for the existence of something like consciousness as a fundamental force or particle since its dependence on being properly assembled from parts to be at all.

Why is whole brain function not greater than the sum of its parts? None of the parts alone are able to produce conscious experience, which is probably one of the greatest things in the World.

 

[Juma]

I understand your feeling about the overuse of "more than the sum of the parts but the brain ofa living person is definitely more than the sum of its parts. When we say "sum of the parts" we think of the parts as in an unorganized heap: simply taking the parts and add them to the heap.