Sign in to follow this  
ahw

memetic and emergence... oh the possibilities!

Recommended Posts

I was gonna answer on his thread, but this is so out of topic I thought I'd start another.
Quote:
by Darkthrone "I, Robot" and "Ghost in the Shell" cites a theory that I don't know if is real-based or not. "In a system, ramdonly protocols that approach that we called "soul"." The Ghost in the Shell theory exists on true? What I find it, if exists?
Ghost in the Shell ! I see you are a dreamer like me [grin] Sadly, dreaming doesn't get things done very quickly, eh? To answer your question, I am sure you'll find the ideas of Richard Dawkins illuminating. His book _The Selfish Gene_ is a great read about genetics and evolution, but where it becomes pure genius is when he explains the concept of memes. Essentially, intelligence is the accidental consequence of genes creating more and more complex organisms to contain and replicate them. Memes are "intellectual genes", if you will. They are ideas, concepts, paradigms, etc, that spread themselves around human cultures (well, I guess animals too could have memes). Things like "Santa Claus wears Red and White" is a pure example of memetic engineering: a company manipulates a well known meme: "Santa Claus"; and cross-breeds it with another: "Coca Cola == Red and White". The result is a new meme that spreads itself on its own, in this case almost entirely erasing the previous one (What are the original colours of Santa Claus costume ? Do you know ?) Once you understand the concept of meme, if you make parallels with genes and their evolution from protozoas to sentient beings, the implications are, well, interesting [grin]. The idea being that once a certain level of complexity has been reached, "something" should emerge out of the primordial memetic soup. Artificial Intelligence, maybe? This is the idea that fuels the emergence of AI in books like Ghost in the Shell. With the chaos that is the Internet, with the amount of memes that are travelling around constantly, what will happen when enough memes aggregate together? Personally I think we are already seeing it happening, but it's nothing as blatant as we expected. Aggregates of memes create cultures, more or less succesfull. It could be the inside jokes of a particular forum (for example, look at the Farkisms of Fark.com), the jargon that evolved across a game (I don't get this Warcraft joke), a genre (everybody that plays FPS knows what "frag","Deathmatch","CTF","ping",etc mean), a medium (the l33tspeak of the Internet...), or more globally, the culture that evolved out of a given civilisation. The question is (if you keep making parallels with organisms), where does a civilisation stands on the evolutionary scale ? And what will happen when the organism reaches self awareness...

Share this post


Link to post
Share on other sites
Memes are an interesting concept. However, to makes memes really work, you need not just be able to pass them on, they must be manipulated and executed before they can contribute to anything.

Since we bring up Ghost in the Shell, I personally feel that with the Ghost in the Shell: Stand Alone Complex series, the AI tanks really bring up interesting issues with AI and where AI can go. The AI tanks were programmed with artificial personalities that started out being thesame, but as each one experiences the world differently, they started becoming slightly different. They completely steal the show and bring up an implicit definition of intelligence and humanity.

That being said, I think intelligence or wisdom or a person is defined by two things, memory (memes) and experience. Two people can be given the same memories, but if their experiences varied, then they become two different people. Thus, we can look at experience as a manipulator or modifier for memes. This has been found true in most recent research in search algorithms, that adaptive searchs with memory mechanisms and a means of manipulating that memory, perform better than those that don't have such mechanics. Of course, I'm not saying that just memory and experience will create intelligence, just that most adaptive AI algorithms, probably should have these two elements.

Share this post


Link to post
Share on other sites
I realise that just rambling about the potential of a particular subject doesn't necessarily lead to an interesting thread, so maybe I should give this some sort of direction by asking a question.

What I am interested in, is exploring the analogy of the meme as a gene.
Essentially, the idea that memes evolve in a manner similar to genes, and seem to aggregate to create new and bigger memes (should we use memotype, to parallel the term "genotype" ?) like jargons, cultural trends, civilisations.

But what I would like to know is, what is the method used to determine the "fitness" of a meme. What are the criteria that our brains use to determine how much they like a meme.

Clearly, memes don't live in the ether, but in the collective soul of Humanity. But of course, this is a figure of speech. The collective soul is, much like the Internet, something that is very much based in hardware, with nodes transmitting and receiving more or less data, with some being much more powerful in their reach than others (an elected governement politician has more reach than your local council politician), and the possibility of interruptions in the flow of data, due to various reasons (I doubt kids in Iraq have heard of the "All your base are belong to us" meme...)

So, the way I see it, each individual brain filters the memes that it receives, and depending on the fitness of a given meme, adopts it and transmits it, ignores it, or possibly fights it by spreading another meme.
For example, some religions quitely ignore others (ignore the new meme), while Single God religions (e.g.) actively fight other religions when exposed to their ideas (for instance, a Muslim shouldn't just ignore somebody talking about their religion, but actually try to show that person the error of their ways; same with Catholic missionaries, who only want to "save souls", help the poor natives see the light...)

Now the question is, how do you know if a new meme is better than another?
What are the mechanisms that make a meme fit in your existing personal memetype.
And why is it that some memes are much more easily changed/overwritten than others?
It seems to me that the more a meme is widespread, the harder it is to change it. Say, "pedophilia is evil" is pretty much global, nowadays, and consequently, any major change to it would occur on the century scale (do I need to remind you that pedophilia was just another sexual practice, in the Classical era?)
On the other hand, a meme like "emo rock", hopefully, should be swallowed up by the next wave of change on the musical scene.

In essence, what I am asking is, I guess, how can we turn the concept of memes from a mind experiment, to something that can be modelled on a computer [grin].
And yes, I am sure we don't need Maths just right now...

Share this post


Link to post
Share on other sites
Quote:
Original post by WeirdoFu
Since we bring up Ghost in the Shell, I personally feel that with the Ghost in the Shell: Stand Alone Complex series, the AI tanks really bring up interesting issues with AI and where AI can go. The AI tanks were programmed with artificial personalities that started out being thesame, but as each one experiences the world differently, they started becoming slightly different. They completely steal the show and bring up an implicit definition of intelligence and humanity.

Like many people who watched the show, I too adore the Tachikoma (even had a little tear for them during the last episode).
Note that they don't have different personalities! What happened is that one of them developed a flaw, which makes it want to explore the world around it. During the daily synchronisation, all Tachikoma share their memories, meaning that they all end up with the memories of the "unique" one.
The joke is that they ALL think that they are the unique one. And of course they all develop different quirks, and opinions about the world. But really, they are just all copies of the one that is starting to become truly sentient (i.e. creative)
In the end, I really liked the remark Kusanagi made, which was that even though they were serialised AI (i.e. clones), and therefore it shouldn't be possible for them to develop a unique personality, there was a solution to this problem: give them curiosity! (and here, I hope you can see the reference to the Bible and the Forbidden Fruit. It's curiosity that started it all!)

Quote:

That being said, I think intelligence or wisdom or a person is defined by two things, memory (memes) and experience.
Two people can be given the same memories, but if their experiences varied, then they become two different people.


I might be missing the point here, but for all intents and purposes, your memories ARE your experience !
If two people have the exact same memories, it doesn't matter what their actual experience was, because it will be forgotten. The problem will be that their will be a gap between the muscular memory and the actual physical state of your muscles (and body, in general). Say, if the original person didn't have arms, and you overwrite someone with arms with his memory, the person with arms won't know what to do with his physically functional arms. I.e. he will have to learn it all over again. (And no, it wouldn't be like waking up with a dead arm and having to "wake it", it would be like being a new born baby, or a like a peron waking up from a ten years coma who has to relearn everything)

Quote:

Thus, we can look at experience as a manipulator or modifier for memes.

That I agree with. A meme that tells you something which you find to be untrue through experience will become less desirable.
Classic examples would be belief in Santa Claus, or in God. Really, it's the same type of meme, with the difference that one is set much much deeper in the collective inconscious that the other. I.e. it's much easier to disbelieve Santa Claus than God, but in both cases, your personal experience will reinforce or lessen your attachment to the given meme. In the case of God, it's easy to say "OK, I don't really believe in the God described in my religion anymore", but it's VERY hard (IMO) to completely remove the meme that "there is something bigger than us"...

I can't comment on adaptive search algorithms [grin]

Share this post


Link to post
Share on other sites
Quote:
Original post by ahw
What I am interested in, is exploring the analogy of the meme as a gene.
Essentially, the idea that memes evolve in a manner similar to genes, and seem to aggregate to create new and bigger memes (should we use memotype, to parallel the term "genotype" ?) like jargons, cultural trends, civilisations.

But what I would like to know is, what is the method used to determine the "fitness" of a meme. What are the criteria that our brains use to determine how much they like a meme.

Clearly, memes don't live in the ether, but in the collective soul of Humanity. But of course, this is a figure of speech. The collective soul is, much like the Internet, something that is very much based in hardware, with nodes transmitting and receiving more or less data, with some being much more powerful in their reach than others (an elected governement politician has more reach than your local council politician), and the possibility of interruptions in the flow of data, due to various reasons (I doubt kids in Iraq have heard of the "All your base are belong to us" meme...)


Interestingly enough, no one ever said that you can't represent memes with genes. Genotype pertains to the actual genetic encoding, while on the representation side, a genotype is mapped to a phenotype, which is like the decoded version of the gene. So, memes are pretty much a phenotype. There was once some speculation that on top of inheriting a parent's physical traits, a child may also inherit personality traits. This brings about, of course, the aruments of lamarckian and baldwinian evolution, which can be an altogether different topic.

But the point is, you can encode memes into genetic form, and there's really no stopping you from doing that, if it isn't already done in nature. Of course, some of it is speculation especially about the nature part, but programming-wise, you can always do that.

Quote:

I might be missing the point here, but for all intents and purposes, your memories ARE your experience !
If two people have the exact same memories, it doesn't matter what their actual experience was, because it will be forgotten. The problem will be that their will be a gap between the muscular memory and the actual physical state of your muscles (and body, in general). Say, if the original person didn't have arms, and you overwrite someone with arms with his memory, the person with arms won't know what to do with his physically functional arms. I.e. he will have to learn it all over again. (And no, it wouldn't be like waking up with a dead arm and having to "wake it", it would be like being a new born baby, or a like a peron waking up from a ten years coma who has to relearn everything)


It is true that experience is part of memory, but i see them as incremental modifiers. The combination of your memories and experiences determine how you experience things now and in the future. You and a friend may see the same car accident happen, but because of your past experiences (which is now part of memory) you may "experience" the incident differently from your friend. The difference in "experience" modifies what you actively remember. So, technically, you both have the same "memory" (saw the same thing, but because of the difference in experience, you two interpret it differently.

I think the confusing part is the fact that experience and memory are usually so intertwined and bound together tightly, that it gets a little hard to differentiate them. For me, I guess its hard to explain it properly. I guess we can say, I may see what you see, but you may not feel what I feel. In the case of the Tachikomas, even though they shared the same information, there's always a slight difference in information storage and interpretation. At least that's how I feel.

Share this post


Link to post
Share on other sites
for those who don't want to read all my rambling...
Quote:

In essence, what I am asking is, I guess, how can we turn the concept of memes from a mind experiment, to something that can be modelled on a computer .

Share this post


Link to post
Share on other sites
Well, I'd say that one of the most important qualities of a meme, is does it seem to give you benefits to believe in it / follow it.

1. Poison kills.
If you drink poison you die. This would be a useful meme, as would don't drink things in unlabelled bottles. Likewise, the meme that you should never put poisons into empty drinks bottles is true because people could drink it and so die.

2. God saves, this is an emotionally beneficial meme, it helps people to think that there is something that looks after them and a purpose to live for.

Of course if a meme, is obviously false then it won't catch on as easily. Ie. Plasticine burns your skin off, which is easily disproved.

In these cases it is a sort of warning of consequences to actions. or a warning not to do things because of unnamed consequences.

Share this post


Link to post
Share on other sites
Quote:
Original post by Ketchaval
1. Poison kills.
If you drink poison you die. This would be a useful meme, as would don't drink things in unlabelled bottles. Likewise, the meme that you should never put poisons into empty drinks bottles is true because people could drink it and so die.


This remind me of something I read on wiki about poisons. Normally the label is a skull and bones, but because of its association with pirates, kids don't really understand it... so they are looking for a better label!
i.e. the meme "skull+bones == pirates" is much stronger than "skull+bones==danger of death" ... at least for kids.

Go figure [wink]

Also, in Ghost in the Shell, the whole Laughing Man logo phenomenon is a big example of a meme that spreads like mad, regardless of its original meaning.

Share this post


Link to post
Share on other sites
Quote:
Original post by ahw
for those who don't want to read all my rambling...
Quote:

In essence, what I am asking is, I guess, how can we turn the concept of memes from a mind experiment, to something that can be modelled on a computer .


Here's a little research I participated in. I think it falls in line with what you want to do as in implementing meme space.

ftp://ftp.eng.auburn.edu/pub/techreports/csse/04/CSSE04-09.pdf

In the research we created a client server set up and have users cooperatively evolve emoticons based on pre-specified criterions. Something like, "an expression that shows irritation." Each user are then presented with 9 possible expressions where they choose the closest one, which is then sent back to the server and placed into meme space. What the users don't know is that within the 9 options they are given, there may be some that came from previous user selections (ones selected by other people as being close). The fitness ranking is implicit and from the server side, you can see how concepts about expressions spread through the general design.

In this scenario, the users collaborate indirectly to create a shared meme about a certain concept. It can also be seen as extracting the "generalized" meme from the various users.

Of course, this research isn't limited to emoticons, it can go alot further.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

Sign in to follow this