After reading the article "Creativity - The last human stronghold?" by Israel Beniaminy over at TFOT, I got inspired to jot down some of my own thoughts on the subject. Let me just emphasize, that even though I might sound like I'm lecturing about the objective truth in the matter, this is only me speculating. To much "I think that..." and "It's my firm belife..." just gets irritating in the long run.
So, what is creativity? What I find especially fascinating is that behaviour, carried out by a machine, seldom is acknowledged as creative. What is it that we have, that the machines are missing?
Would you trust this guy to repaint your bedroom?
Let's first ask ourselfs: Is creativity present under circumstances we normaly don't associate with this skill?
Imagine you're searching for something in a limited space. Let's say you're trying to locate your keys which are somewhere in your house. How do you proceed?
Strictly speaking, there are more places where they could be than you could posibly cover in a somewhat reasonable time. However, some places are meaningless to even consider. You won't look in the shower or in the dishwasher, simply because you never would put your keys there. Many options are in other words excluded from the begining. Of course, you could ask yourself the question whether these alternatives is processed by the brain at all. In the same way as you don't consider traveling to another country to try to retrive your keys, you probably don't either take the bathroom alternative into account. Some other, more plausible options, may however be taken into consideration and rejected.
Anyhow, you could attack the problem using a lot of different strategies, but you'll probably make some kind of mental list over posible places. This list is probably hierarchic organized too. You look in your jack pockets before you start to get really desperate and search through your loundary basket. Sometimes you'll get an spontaneous idea just by being in a certain situation. You might for instance look in the top drawer when you've just gone through the one on the bottom. Not because you had that intention when you walked up to the bureau, but because the situation generated that idea.
Has this whole search process something to do with creativity? Well. Let's leave these infernal keys for a moment and go to an other area which is highly accosiated with creativity. Let's go to another set of keys. The piano.
What happens when you sit down by the piano, ready to compose a masterpiece? Well...If it's your first time playing, chances are that what you accomplish probably will sound like crap. But let us say that you have some experience.
First of all, like in the last example, there are a lot of key combinations that you simply even won't consider. You have a sort of repertoire with acceptable moves and combinations. The first thing you play, if you don't really have a plan of what your going to create, will probably be something random. However, from there, there are a lot of options that will make your ears hurt. As soon as you've started playing, your options are limited and you have to apply your, either implicit or explicit, knowledge of scales, harmony, counterpoint and so on. However, it isn't a strict formal process that always will lead to the same result. The outcome depends on your current mood, what you focus on in the music, your habits, what you happend to associate with what according to certain circumstances and so on. Everything is connected in an extremely complicated network of calculations, and the slightest change of a parameter can have profound impact on the outcome.
Compare this situation to something simmilar (I will try to tie up all these examples in the end. Bare with me). Imagine you're producing linguistic statements. The statements doesn't really have to make sense. The only thing that matters is that you keep producing them, even if you really can't think of any meaningfull content to put into them. For example, this is what I produced right now: "I wan't to think that then it isn't really that bad but couldn't you say that again if that's no big problem." Even though I probably would be institutionalized if this was how I usually talked, you can clearly see somewhat of a pattern here. If you read the the whole sentence it's just confusing, but by looking at segments of it, it's actually coherent. "I want to think that", "it isn't really that bad" and "no big problem" are all fully understandable and usable segments that can be put into meaningfull sentences.
Could one argue that this situation is similar to the piano example? Clearly you're following a given set of rules. You have a grammar that you really don't violate and exactly which words you choose to utter depends on a lot of different factors that togheter decide the outcome. So, are both the latter and the former situations examples of someone using their creativity? That naturally depends on how you choose to define the concept of creativity, but let me try to distinguish what's really differs beetween the two.
A computer could carry out the latter task, producing linguistic sentences, without any bigger problems. You just have to give the program at hand a couple of words, assign them to different clause elements and then put down some rules of how it's allowed to combine the different grammatic parts. It won't be perfect, and noone has so far managed to make a fully explicit description of a gramatic system in any human language, but it would be fairly easy to put something togheter that performed about as good as a real human. Is this an example of creativity? Well, some people maybe say so, but it really just boils down to a given set of rules and a couple of randomly asigned values. Not so exciting.
Well, let's turn to the other example. Playing the piano. Isn't this really just a set of patterns, rules and learned behaviours? Well, yes it is. But what makes this behaviour special is all in the feedbak process.
Think about the key example again. What kind of feedback process does take place in it? Well, nothing really. It's pretty simple. If you find the keys, you stop looking. If you don't, you continue the search. There isn't any kind of situation where you sort of find them. So, the feedback response is binary and purely used to decide whether the behaviour should be canceled or not.
Let's get back to the piano. Now, things get a little more complicated. There isn't one way to play it that is the right way. The performance is constantly being judged by the brain on the basis of a lot of different factors. Rythm, harmony, tension, mood, constancy...The list could go on for a pretty long time. If something sounds good, we reinforce the behaviour that led to it. If it sounds bad, we repress it. Of course, this process is at work when we search for our keys too, but to a lesser extent.
I suggest this is the key (there are an awful lot of keys in this text) to creativity. The reason we haven't been able to successfully inplant this preciouss behaviour into computers, is because we haven't figured out what makes us tick, when it comes to these areas accociated with creativity. It's a complicated process to create a computer program that does something we want. It's even harder to make it do it in the same way we do.
I would like to say that computers, in theory, can be just as creative as us humans. The problem is that they're working with another set of preferences than us. Of course we are going to look at the result and say it's worthless when it's created by something that doesn't share our views on what's good. If you would let a cat compose its own meal (don't ask me how that could be carried out) we probably wouldn't like that either. That doesn't mean that the cat did something wrong according to its own measures. It just means that we don't like tuna that much.
So, what is creativity? What I find especially fascinating is that behaviour, carried out by a machine, seldom is acknowledged as creative. What is it that we have, that the machines are missing?
Would you trust this guy to repaint your bedroom?
Let's first ask ourselfs: Is creativity present under circumstances we normaly don't associate with this skill?
Imagine you're searching for something in a limited space. Let's say you're trying to locate your keys which are somewhere in your house. How do you proceed?
Strictly speaking, there are more places where they could be than you could posibly cover in a somewhat reasonable time. However, some places are meaningless to even consider. You won't look in the shower or in the dishwasher, simply because you never would put your keys there. Many options are in other words excluded from the begining. Of course, you could ask yourself the question whether these alternatives is processed by the brain at all. In the same way as you don't consider traveling to another country to try to retrive your keys, you probably don't either take the bathroom alternative into account. Some other, more plausible options, may however be taken into consideration and rejected.
Anyhow, you could attack the problem using a lot of different strategies, but you'll probably make some kind of mental list over posible places. This list is probably hierarchic organized too. You look in your jack pockets before you start to get really desperate and search through your loundary basket. Sometimes you'll get an spontaneous idea just by being in a certain situation. You might for instance look in the top drawer when you've just gone through the one on the bottom. Not because you had that intention when you walked up to the bureau, but because the situation generated that idea.
Has this whole search process something to do with creativity? Well. Let's leave these infernal keys for a moment and go to an other area which is highly accosiated with creativity. Let's go to another set of keys. The piano.
What happens when you sit down by the piano, ready to compose a masterpiece? Well...If it's your first time playing, chances are that what you accomplish probably will sound like crap. But let us say that you have some experience.
First of all, like in the last example, there are a lot of key combinations that you simply even won't consider. You have a sort of repertoire with acceptable moves and combinations. The first thing you play, if you don't really have a plan of what your going to create, will probably be something random. However, from there, there are a lot of options that will make your ears hurt. As soon as you've started playing, your options are limited and you have to apply your, either implicit or explicit, knowledge of scales, harmony, counterpoint and so on. However, it isn't a strict formal process that always will lead to the same result. The outcome depends on your current mood, what you focus on in the music, your habits, what you happend to associate with what according to certain circumstances and so on. Everything is connected in an extremely complicated network of calculations, and the slightest change of a parameter can have profound impact on the outcome.
Compare this situation to something simmilar (I will try to tie up all these examples in the end. Bare with me). Imagine you're producing linguistic statements. The statements doesn't really have to make sense. The only thing that matters is that you keep producing them, even if you really can't think of any meaningfull content to put into them. For example, this is what I produced right now: "I wan't to think that then it isn't really that bad but couldn't you say that again if that's no big problem." Even though I probably would be institutionalized if this was how I usually talked, you can clearly see somewhat of a pattern here. If you read the the whole sentence it's just confusing, but by looking at segments of it, it's actually coherent. "I want to think that", "it isn't really that bad" and "no big problem" are all fully understandable and usable segments that can be put into meaningfull sentences.
Could one argue that this situation is similar to the piano example? Clearly you're following a given set of rules. You have a grammar that you really don't violate and exactly which words you choose to utter depends on a lot of different factors that togheter decide the outcome. So, are both the latter and the former situations examples of someone using their creativity? That naturally depends on how you choose to define the concept of creativity, but let me try to distinguish what's really differs beetween the two.
A computer could carry out the latter task, producing linguistic sentences, without any bigger problems. You just have to give the program at hand a couple of words, assign them to different clause elements and then put down some rules of how it's allowed to combine the different grammatic parts. It won't be perfect, and noone has so far managed to make a fully explicit description of a gramatic system in any human language, but it would be fairly easy to put something togheter that performed about as good as a real human. Is this an example of creativity? Well, some people maybe say so, but it really just boils down to a given set of rules and a couple of randomly asigned values. Not so exciting.
Well, let's turn to the other example. Playing the piano. Isn't this really just a set of patterns, rules and learned behaviours? Well, yes it is. But what makes this behaviour special is all in the feedbak process.
Think about the key example again. What kind of feedback process does take place in it? Well, nothing really. It's pretty simple. If you find the keys, you stop looking. If you don't, you continue the search. There isn't any kind of situation where you sort of find them. So, the feedback response is binary and purely used to decide whether the behaviour should be canceled or not.
Let's get back to the piano. Now, things get a little more complicated. There isn't one way to play it that is the right way. The performance is constantly being judged by the brain on the basis of a lot of different factors. Rythm, harmony, tension, mood, constancy...The list could go on for a pretty long time. If something sounds good, we reinforce the behaviour that led to it. If it sounds bad, we repress it. Of course, this process is at work when we search for our keys too, but to a lesser extent.
I suggest this is the key (there are an awful lot of keys in this text) to creativity. The reason we haven't been able to successfully inplant this preciouss behaviour into computers, is because we haven't figured out what makes us tick, when it comes to these areas accociated with creativity. It's a complicated process to create a computer program that does something we want. It's even harder to make it do it in the same way we do.
I would like to say that computers, in theory, can be just as creative as us humans. The problem is that they're working with another set of preferences than us. Of course we are going to look at the result and say it's worthless when it's created by something that doesn't share our views on what's good. If you would let a cat compose its own meal (don't ask me how that could be carried out) we probably wouldn't like that either. That doesn't mean that the cat did something wrong according to its own measures. It just means that we don't like tuna that much.
No comments:
Post a Comment