Creative thinking - the little we know.
Here is where we must once again delve into philosophy. We must use our critical faculties to sift what is known, from what is merely hypothesized, or hoped for, uncoupled from reality. Here is a rule of thumb (again, due to David Deutsch):
If we understand something, we can program it. Allow me a moment to unpack that. Here are some things we understand to greater or lesser degrees:
We understand simple arithmetic. We know how to add numbers, subtract them - we can divide and do things like factorise them. Indeed we know we understand all this because we can program and then build electronic calculators that do all these things. We have understood much of this for millenia.
We understand Newton’s Laws of Motion and his theory of gravity. We can program a computer with virtual solar systems and with those calculate where a planet will be at any given time from now to a thousand years hence. The formulae that allow us to do this have been known for centuries.
We have some understanding of chemical reactions, and biological natural selection - we can program computers with these things too and make predictions.
Here is something we understand far less well: how economic markets work. This is not to say we understand exactly nothing - but we know far less than most people seem to think. We cannot predict the behaviour of markets well at all. And we actually have good explanations as to why such prediction is beyond us: the complexity is too great for us to capture with any simple mathematical theory that could be represented in a computer system. In short: there are too many finely tuned variables and we don’t even know all of the variables that come to bear on how markets react to changes in, for example, fluctuations in the value of the dollar because we do not even understand how fluctuations in the dollar are affected by (well now name yet another variable that affects changes in the dollar). The multitudinous feedback mechanisms at work here are explained in a very approximate way by economic theories that themselves, if they are at all any good, explain why predicting markets is, for now, completely infeasible. And might actually be intractable forever. (This is because a truly predictive model of a market would have to account for the free choices people make. And that would mean predicting what a person does. Despite what some economists, sociologists, psychologists, philosophers or neuroscientists might attempt to imply: this is something we cannot do - nor should we desire to).
Now here is something we not only know little about - but one is tempted to say we know next-to-nothing about: creativity. Human Creativity is a truly unique feature of the world as we know it in the year I write this (2015). Human creativity is not the only type of creativity. There is at least one other I am aware of: evolution by natural selection. Evolution by natural selection create new species. That is a form of creativity. But it is not conscious - it is not thought through.
Let us call that kind of creativity: biological creativity. That kind of creativity is a physical process where genes are rejected by the environment in which they find themselves. If a gene “works” it is kept (by an organism) until it doesn’t. And the organism dies and if that happens enough, the gene - along with its host species - goes extinct. Genes are the unit of selection. And nature is what criticises. Genes might survive the criticism. Or not. A criticism of a gene might amount to the environment changing in such a way as to restrict water: that is to say a drought happens. In this case - the gene may not survive. Variants of the gene that code for (say) the capacity to survive with less water will survive because the organism survives. In this way, variation within a species leads to speciation: a new species more “drought resistant” than the one before the drought survives.
I do not wish to push an analogy too far here: but Human Creativity and the Growth of Knowledge approximately resembles this process. A new idea is like a new gene - it is typically a variation on an old idea (as a mutation is a variation of an older version of a gene). Most gene variations (mutuations) are bad. Most new ideas are bad. But now and again a variant is actually better. It survives the criticisms. In biology this process is all automated - it is simply an outworking of the interactions between unthinking genes and an unthinking environment.
But in people things are quite different. The basic idea is the same: new things are criticised. But the means by which those new things come about at all is very different. New genes are mututations - and we understand how mutuations occur: it is bad copying. It could be something as prosaic as a cosmic ray (a particle of light) which hits the DNA and changes it.
How does human creativity work? Here is what we do know:
Creativity is about making a variation to an existing idea. One cannot wholesale invent ex-nihilio. At a minimum the idea has to at least be expressible in some form: natural language, paint, a set of numbers. So there has to be some pre-existing human creation there for one to adapt. But then what?
We don’t know.
And that’s the truth. We do not know. If we knew how it was we created new ideas we could program a computer to do it. And this, by the way, is why we do not have computers that can think. This is why we do not have artificial intelligence (or artificial general intelligence AGI). The problem is not the hardware - the physical stuff the computer is made from. Right now silicon valley is ablaze with the idea that somehow, via Moore’s Law (the idea that computer power, speed and memory is doubling every 18 months) is somehow going to precipitate AGI. As though faster computers that can store more stuff - or the internet itself - is about to come alive and start learning. As though it is just a matter of speed and memory. This is completely false. It’s not just wide of the mark - it’s not even shooting for the right target.
The problem has nothing to do with speed or memory. That is, it has nothing whatever to do with hardware. It has everything to do with software. With the program. We have no program of “thinking” - which is to say we have no way of programming genuine human creativity. We don’t know how it works - unlike with arithmetic, or the laws of physics or even the theory of natural selection. We can program those things to get a computer to simulate them. But not thinking. Because we don’t understand it.
And here’s the thing: we have to understand those things before we program computers to do it.
It is not the case that one day we will find a thinking computer and then wonder: how is it doing that? No. First we will have an explanation of how thinking works in people and then we will know how to program a computer to think.
So here is the corollary of that: because we do not understand how new ideas are actually generated (which is to say, we don’t understand how people think) and we know this because we cannot program computers to think - we cannot hope to improve the creativity of people. Because we do not have a theory of creativity that works. Even approximately.
I have sketched here the basics of what we do know about creativity: namely - next to nothing. We know creativity at a minimum involves variations on ideas already there. But beyond that - we know very little else.
Now all humans are creative. That is what humans truly are. Creative beings. Insofar as other animals demonstrate creativity - it is orders of magnitude lesser. So consider our creative luminaries:
Alan Turing: who invented the theory of the computer which eventually led to the building of one.
Albert Einstein: who explained the general theory of relativity allowing for the global positioning system and (therefore) Google maps to help you find the nearest cafe.
Salvidor Dali: whose melting clocks gave us a surrealist and disturbing insight into our experience of time.
Amadeus Mozart: almost the Einstein of music - whose technical versatility was able to conjure mere sound in strange new ways forming deeply emotional music unlike much that had come before.
If we could explain how any of this was actually done - we could reproduce a computer to do it. Or teach a child to do it. But we cannot. A child teaches us what they can do just as Turing, Einstein, Dali and Mozart taught us. We do not teach them. And could you imagine daring to suggest to any of these masters that we possessed techniques to encourage creative thinking?
What would they have needed to improve their creativity? Here are two things: freedom and time. That is all. Freedom to explore any idea, technique or method they liked. Freedom to explore those ideas when and where they pleased. Freedom to do what they were most interested and passionate about. And time to do this. Whenever they needed it.
Now further - imagine you said to Einstein: to be well rounded you need to spend this time now playing Mozart and this time now learning poetry and this time now learning to speak Cantonese. The time taken to do this? Maybe we would not have GPS now.
Imagine forcing Mozart to learn more algebra. Or chemistry.
Would this help their creativity - or not? The point here is not that those subjects would be useless for those people it is that we would be choosing for them. Einstein did actually value his violin almost above physics. But he chose where to play it and when.
Iterative cycles of creative (undirected adaptions of existing knowledge) coupled with critical (criticisms of those undirected adaptations in order to trim the false from the closer-to-true) thinking is the way learning occurs. This is what we know. It is important to emphasise that this is a process that goes on inside your mind and is not a mere physical activity of neurones (although it is also that is not primarily that). This is why learning is not within the purview of neuroscience anymore than art is within the purview of physics. Let me explain: I know some physics. I know some physics about light and colour. I can explain something of the interaction between matter and energy and how different substances, when excited by energy (be in light shining upon it, or thermal energy heating it) it might appear to be green or red or blue or some combination thereof. I know about these processes. That's physics. But just because I can explain some of that, does not mean I know anything at all about the "light" and "shade" used in art. The two ways "colour" are used in the two subjects have something to do with each other - of course. But expertise in one does not necessarily constitute expertise in the other. My understanding of the physics of light does not in any way enable me to be able to explain how light should be used in a painting in order to either make it look more realistic or to convey a certain mood or feeling. That is a whole different type of knowledge - at a different "level of emergence" and complexity.
So it is here. We hear much about how so-called "brain science" might inform "learning". But this is overblown. We never learn much from MRI scans of the brain or what neuroscience has learned about some new structure identified in the hippocampus. We never will. For the same reason an artist, expert in painting, is not going to learn much - ever - about how colour is generated by the interaction between photons of light and electrons around atoms. It's the wrong level of emergence. The wrong type of complexity. To be more specific the artist is interested in how paint can abstractly represent other physical objects. The physicist is concerned solely with the physical objects in themselves.
So it is here. The epistemologist (or psychologist) interested in learning should be concerned with how a person is able to take pieces of knowledge and (re-)create it in their own mind through the process of creative and critical thinking. At no point do we need to refer to neurones, neurotransmitters, the pre-frontal cortex and synapses - that is - the physical structures in which the knowledge is embedded. Not in order to understand learning. It's not that it's entirely irrelevant (for now, wet brains are the sole objects in the universe that we know of that can actually create new explanatory knowledge) but to think it is of fundamental importance that the hardware is a certain way is in the wrong direction entirely. It really is like thinking that if you're having trouble using your word processing program, that a better understanding of the silicon chips inside the computer just might be what's required. The study of silicon chips is, of course, crucial to understanding computers as a whole - but not all problems with computers are reducible to silicon chips! And so it very much is with minds. Not all our ignorance (for example about how learning works best) when it coems to the mind is reducible to what aspects of the brain like neurones, neurotransmitters, synapses and brain regions are doing. Indeed only a tiny minority of ways in which one might improve ones own mind will be usefully informed, in this way, by neuroscience. The distinction here really is one of software versus hardware. Confusing the two leads to much confusion.
Final thoughts.
If we understand something, we can program it. Allow me a moment to unpack that. Here are some things we understand to greater or lesser degrees:
We understand simple arithmetic. We know how to add numbers, subtract them - we can divide and do things like factorise them. Indeed we know we understand all this because we can program and then build electronic calculators that do all these things. We have understood much of this for millenia.
We understand Newton’s Laws of Motion and his theory of gravity. We can program a computer with virtual solar systems and with those calculate where a planet will be at any given time from now to a thousand years hence. The formulae that allow us to do this have been known for centuries.
We have some understanding of chemical reactions, and biological natural selection - we can program computers with these things too and make predictions.
Here is something we understand far less well: how economic markets work. This is not to say we understand exactly nothing - but we know far less than most people seem to think. We cannot predict the behaviour of markets well at all. And we actually have good explanations as to why such prediction is beyond us: the complexity is too great for us to capture with any simple mathematical theory that could be represented in a computer system. In short: there are too many finely tuned variables and we don’t even know all of the variables that come to bear on how markets react to changes in, for example, fluctuations in the value of the dollar because we do not even understand how fluctuations in the dollar are affected by (well now name yet another variable that affects changes in the dollar). The multitudinous feedback mechanisms at work here are explained in a very approximate way by economic theories that themselves, if they are at all any good, explain why predicting markets is, for now, completely infeasible. And might actually be intractable forever. (This is because a truly predictive model of a market would have to account for the free choices people make. And that would mean predicting what a person does. Despite what some economists, sociologists, psychologists, philosophers or neuroscientists might attempt to imply: this is something we cannot do - nor should we desire to).
Now here is something we not only know little about - but one is tempted to say we know next-to-nothing about: creativity. Human Creativity is a truly unique feature of the world as we know it in the year I write this (2015). Human creativity is not the only type of creativity. There is at least one other I am aware of: evolution by natural selection. Evolution by natural selection create new species. That is a form of creativity. But it is not conscious - it is not thought through.
Let us call that kind of creativity: biological creativity. That kind of creativity is a physical process where genes are rejected by the environment in which they find themselves. If a gene “works” it is kept (by an organism) until it doesn’t. And the organism dies and if that happens enough, the gene - along with its host species - goes extinct. Genes are the unit of selection. And nature is what criticises. Genes might survive the criticism. Or not. A criticism of a gene might amount to the environment changing in such a way as to restrict water: that is to say a drought happens. In this case - the gene may not survive. Variants of the gene that code for (say) the capacity to survive with less water will survive because the organism survives. In this way, variation within a species leads to speciation: a new species more “drought resistant” than the one before the drought survives.
I do not wish to push an analogy too far here: but Human Creativity and the Growth of Knowledge approximately resembles this process. A new idea is like a new gene - it is typically a variation on an old idea (as a mutation is a variation of an older version of a gene). Most gene variations (mutuations) are bad. Most new ideas are bad. But now and again a variant is actually better. It survives the criticisms. In biology this process is all automated - it is simply an outworking of the interactions between unthinking genes and an unthinking environment.
But in people things are quite different. The basic idea is the same: new things are criticised. But the means by which those new things come about at all is very different. New genes are mututations - and we understand how mutuations occur: it is bad copying. It could be something as prosaic as a cosmic ray (a particle of light) which hits the DNA and changes it.
How does human creativity work? Here is what we do know:
Creativity is about making a variation to an existing idea. One cannot wholesale invent ex-nihilio. At a minimum the idea has to at least be expressible in some form: natural language, paint, a set of numbers. So there has to be some pre-existing human creation there for one to adapt. But then what?
We don’t know.
And that’s the truth. We do not know. If we knew how it was we created new ideas we could program a computer to do it. And this, by the way, is why we do not have computers that can think. This is why we do not have artificial intelligence (or artificial general intelligence AGI). The problem is not the hardware - the physical stuff the computer is made from. Right now silicon valley is ablaze with the idea that somehow, via Moore’s Law (the idea that computer power, speed and memory is doubling every 18 months) is somehow going to precipitate AGI. As though faster computers that can store more stuff - or the internet itself - is about to come alive and start learning. As though it is just a matter of speed and memory. This is completely false. It’s not just wide of the mark - it’s not even shooting for the right target.
The problem has nothing to do with speed or memory. That is, it has nothing whatever to do with hardware. It has everything to do with software. With the program. We have no program of “thinking” - which is to say we have no way of programming genuine human creativity. We don’t know how it works - unlike with arithmetic, or the laws of physics or even the theory of natural selection. We can program those things to get a computer to simulate them. But not thinking. Because we don’t understand it.
And here’s the thing: we have to understand those things before we program computers to do it.
It is not the case that one day we will find a thinking computer and then wonder: how is it doing that? No. First we will have an explanation of how thinking works in people and then we will know how to program a computer to think.
So here is the corollary of that: because we do not understand how new ideas are actually generated (which is to say, we don’t understand how people think) and we know this because we cannot program computers to think - we cannot hope to improve the creativity of people. Because we do not have a theory of creativity that works. Even approximately.
I have sketched here the basics of what we do know about creativity: namely - next to nothing. We know creativity at a minimum involves variations on ideas already there. But beyond that - we know very little else.
Now all humans are creative. That is what humans truly are. Creative beings. Insofar as other animals demonstrate creativity - it is orders of magnitude lesser. So consider our creative luminaries:
Alan Turing: who invented the theory of the computer which eventually led to the building of one.
Albert Einstein: who explained the general theory of relativity allowing for the global positioning system and (therefore) Google maps to help you find the nearest cafe.
Salvidor Dali: whose melting clocks gave us a surrealist and disturbing insight into our experience of time.
Amadeus Mozart: almost the Einstein of music - whose technical versatility was able to conjure mere sound in strange new ways forming deeply emotional music unlike much that had come before.
If we could explain how any of this was actually done - we could reproduce a computer to do it. Or teach a child to do it. But we cannot. A child teaches us what they can do just as Turing, Einstein, Dali and Mozart taught us. We do not teach them. And could you imagine daring to suggest to any of these masters that we possessed techniques to encourage creative thinking?
What would they have needed to improve their creativity? Here are two things: freedom and time. That is all. Freedom to explore any idea, technique or method they liked. Freedom to explore those ideas when and where they pleased. Freedom to do what they were most interested and passionate about. And time to do this. Whenever they needed it.
Now further - imagine you said to Einstein: to be well rounded you need to spend this time now playing Mozart and this time now learning poetry and this time now learning to speak Cantonese. The time taken to do this? Maybe we would not have GPS now.
Imagine forcing Mozart to learn more algebra. Or chemistry.
Would this help their creativity - or not? The point here is not that those subjects would be useless for those people it is that we would be choosing for them. Einstein did actually value his violin almost above physics. But he chose where to play it and when.
Iterative cycles of creative (undirected adaptions of existing knowledge) coupled with critical (criticisms of those undirected adaptations in order to trim the false from the closer-to-true) thinking is the way learning occurs. This is what we know. It is important to emphasise that this is a process that goes on inside your mind and is not a mere physical activity of neurones (although it is also that is not primarily that). This is why learning is not within the purview of neuroscience anymore than art is within the purview of physics. Let me explain: I know some physics. I know some physics about light and colour. I can explain something of the interaction between matter and energy and how different substances, when excited by energy (be in light shining upon it, or thermal energy heating it) it might appear to be green or red or blue or some combination thereof. I know about these processes. That's physics. But just because I can explain some of that, does not mean I know anything at all about the "light" and "shade" used in art. The two ways "colour" are used in the two subjects have something to do with each other - of course. But expertise in one does not necessarily constitute expertise in the other. My understanding of the physics of light does not in any way enable me to be able to explain how light should be used in a painting in order to either make it look more realistic or to convey a certain mood or feeling. That is a whole different type of knowledge - at a different "level of emergence" and complexity.
So it is here. We hear much about how so-called "brain science" might inform "learning". But this is overblown. We never learn much from MRI scans of the brain or what neuroscience has learned about some new structure identified in the hippocampus. We never will. For the same reason an artist, expert in painting, is not going to learn much - ever - about how colour is generated by the interaction between photons of light and electrons around atoms. It's the wrong level of emergence. The wrong type of complexity. To be more specific the artist is interested in how paint can abstractly represent other physical objects. The physicist is concerned solely with the physical objects in themselves.
So it is here. The epistemologist (or psychologist) interested in learning should be concerned with how a person is able to take pieces of knowledge and (re-)create it in their own mind through the process of creative and critical thinking. At no point do we need to refer to neurones, neurotransmitters, the pre-frontal cortex and synapses - that is - the physical structures in which the knowledge is embedded. Not in order to understand learning. It's not that it's entirely irrelevant (for now, wet brains are the sole objects in the universe that we know of that can actually create new explanatory knowledge) but to think it is of fundamental importance that the hardware is a certain way is in the wrong direction entirely. It really is like thinking that if you're having trouble using your word processing program, that a better understanding of the silicon chips inside the computer just might be what's required. The study of silicon chips is, of course, crucial to understanding computers as a whole - but not all problems with computers are reducible to silicon chips! And so it very much is with minds. Not all our ignorance (for example about how learning works best) when it coems to the mind is reducible to what aspects of the brain like neurones, neurotransmitters, synapses and brain regions are doing. Indeed only a tiny minority of ways in which one might improve ones own mind will be usefully informed, in this way, by neuroscience. The distinction here really is one of software versus hardware. Confusing the two leads to much confusion.
Final thoughts.