On Souls and Fourier Transforms
Do algorithms have souls?
On the surface this is an absurd question - of course not. How could they? An algorithm is just a set of instructions; sometimes simple (like a traffic light's flashing pattern), sometimes complex (like the math powering Google). So let me phrase the question another way:
Do Human Beings have souls? (For the sake of argument I'll assume you answer yes, atheists can skip to the next bit)
A Human in a coma?
Functionally brain dead Human bodies kept alive by machines?
Chimpanzees and Gorillas?
Dogs?
Mice?
Grasshoppers?
Fleas?
Bacteria?
Viruses?
The simplest viruses are on the order of 100 proteins long; can 100 proteins linked in a chain have a soul? How about the same virus, but created artificially by combining 100 pre-prepared proteins? If so, does a computer model of that virus, accurate to the atomic level, also own a soul?
For the nonreligious, you can rephrase the same inquiry in terms of consciousness - the question is just as difficult.
These debates have a long history - religious arguments about whether man is "Separate from the beasts and steward of the Earth" go back thousands of years. The simplest answer is typically "People have souls, nothing else does. Brain function is irrelevant - we are separate from the animals." Unfortunately, this answer may become insufficient within our lifetimes as we struggle to grasp the ethical ramifications of Artificial Intelligence. In my mind, the only logically consistent way to attribute consciousness to a protein chain or a computer program is a kind of scientific animism. I'm a nerd, so I like math analogies. You may have heard of a mathematical function called a Fourier Transform; engineers like to talk about them a lot. I don't want to dig into the math and bore you with details, so here's the simple explanation: a Fourier transform converts information from the world of time into the world of frequencies. There's some calculus involved so that the conversion can be perfect; the transformation requires negative and positive infinities in order to make sense. The astonishing part is that the conversion can be exact - every imaginable piece of information (from a light beam carrying color to an indie rock concert to a single tree falling in a forest) can be expressed in the frequency domain exactly; the inverse transform can take a frequency description and exactly express it in the time domain (from negative infinite time to positive infinite time).
So why bother? What good is a Fourier Transform? There's no way to touch, see or perceive the frequency domain; it's an intangible concept. If you were to walk onto a college campus and argue about the tangible existence of the frequency domain, students would try to buy drugs from you and professors would laugh in your face - that's not a useful question because it's not something that can be known in our universe. The utility of a Fourier Transform lies in approximation - it can be "boiled down" from infinity to a usefully small amount of information and time, enough that the approximations can solve some quite interesting problems. Alternating Current (AC) electrical systems, guitar effects pedals/tuners, photoshop and missile guidance systems all make use of Fourier Analysis. Autotune (the reason all pop music sounds the same now) was originally developed as a Fourier Analysis program for oil mining.
By analogy, debating about the tangible existence of a soul (or the tangible existence of consciousness) is an endlessly circular debate because it doesn't have an answer. Our world of time, space, atoms and chemistry doesn't have room for a soul separate from a body, for consciousness without a brain. However, just because something is intangible doesn't mean it can't exist; just that the question of existence is poorly phrased (because it doesn't specify which universe you mean). In an analogy to Fourier transforms converting from our world of time to an imaginary realm of pure frequencies, I imagine a conversion from our world of atoms to an imaginary realm of consciousness (similar to Plato's realm of ideals). The idea has some appeal to it - every algorithm, every virus, every atom has some representation in this "thought domain". Every possible consciousness - every person who ever was, will be or might have been all exists in an intangible, invisible universe with no concept of time or space. When you die in the realm of atoms, your body may cease to function - but in the intangible thought domain your consciousness lives, along with every possible variation of your consciousness, your past consciousness from 5 years ago, the consciousness of just your amygdala on January 24th, 2005, your grandmother's consciousness before her stroke, the consciousness of a bacteria living on your left pinky toe, the consciousness of an app on your phone; everything. If you grew up religious you might recognize this description as a rephrasing of some very old concepts - to use some more math analogies, the total summation of everything in the thought domain into a single unified whole is sometimes referred to as "God". Rephrased, the concept of "God" involves the entirety of time, space, our universe and everything in it all functioning as a single unified consciousness of which we are all parts. Neat idea, impossible to prove right or wrong without resorting to circular logic.
So what good is this "thought domain"? How is it helpful? Also by analogy with the Fourier Transform - the important questions don't invoke existence, they are focused on utility. Colloquial usage for "does a creature have a soul?" is really better phrased "is it ok for me to abuse this creature?". Asking whether a particular collection of atoms deserves empathy and compassion - whether it qualifies as "having a soul" or "not having a soul" - is like asking whether a collection of frequencies qualifies as music. The answer is entirely subjective; there can be no objective definition of what music means (except Nickelback. I think we can all agree Nickelback doesn't count as music.)
Is it ok to kill a virus? Yes.
Is it acceptable to kill a flea? Yeah, that's fine.
How about a mouse? (Actually don't answer that, I might lose some friends on that question)
Historically, the evaluation of who "gets" a soul and who doesn't (and is therefore acceptable to abuse) was typically tied into political and economic factors. My prediction is that AI will only "get" a soul when it reaches sufficient complexity to fight for classification as such (think Black Mirror-esque brain digitizations). Like a Turing test, this is a messy and subjective way to do things, but I can't see a viable alternative.
So, as with the Fourier Transform, the really important question when thinking about consciousness and souls isn't "Do they exist?", it's "How is this definition useful?". If you believe in souls, religion, God, all that stuff - how does it make you act? Do you use that information to forge bonds with those unlike yourself; to seek commonalities and heal divisions wherever possible?
I've seen firsthand the power of religion to forge bonds, unite communities and strengthen those in need of help. Pretty much every religion focuses on Peace and Unity (the Arabic greeting "Salaam Alaikum" literally translates to "Peace Be Unto You" - the same phrase used extensively in Catholic services). However, every religion is a human institution; each has lots of biases and historical bullshit built into it (if that statement offends you, you can be assured that I'm not talking about YOUR religion. Yours is absolutely true and perfect. I'm talking about someone else's.)
Atheists can argue at length about the (non)existence of intangibles; again, I don't find the existence question meaningful. The more important question is utility - what does your atheism cause you to do? Does it mean that you spend time helping those who historically were crushed under the wheels of organized religion? Does it motivate you to show people that secular humanism can be just as much a powerful force of good as an organized religion? Alternatively, does it cause you to sink into apathy, isolation and nihilism?
Regardless of what beliefs you hold - do they make the world a better place? Do they make you a better person or a worse person? If your belief in Peace is exploited in order to justify hatred, then is it actually a deeply-held conviction or just words you say in order to fit into a community?
On the surface this is an absurd question - of course not. How could they? An algorithm is just a set of instructions; sometimes simple (like a traffic light's flashing pattern), sometimes complex (like the math powering Google). So let me phrase the question another way:
Do Human Beings have souls? (For the sake of argument I'll assume you answer yes, atheists can skip to the next bit)
A Human in a coma?
Functionally brain dead Human bodies kept alive by machines?
Chimpanzees and Gorillas?
Dogs?
Mice?
Grasshoppers?
Fleas?
Bacteria?
Viruses?
The simplest viruses are on the order of 100 proteins long; can 100 proteins linked in a chain have a soul? How about the same virus, but created artificially by combining 100 pre-prepared proteins? If so, does a computer model of that virus, accurate to the atomic level, also own a soul?
For the nonreligious, you can rephrase the same inquiry in terms of consciousness - the question is just as difficult.
These debates have a long history - religious arguments about whether man is "Separate from the beasts and steward of the Earth" go back thousands of years. The simplest answer is typically "People have souls, nothing else does. Brain function is irrelevant - we are separate from the animals." Unfortunately, this answer may become insufficient within our lifetimes as we struggle to grasp the ethical ramifications of Artificial Intelligence. In my mind, the only logically consistent way to attribute consciousness to a protein chain or a computer program is a kind of scientific animism. I'm a nerd, so I like math analogies. You may have heard of a mathematical function called a Fourier Transform; engineers like to talk about them a lot. I don't want to dig into the math and bore you with details, so here's the simple explanation: a Fourier transform converts information from the world of time into the world of frequencies. There's some calculus involved so that the conversion can be perfect; the transformation requires negative and positive infinities in order to make sense. The astonishing part is that the conversion can be exact - every imaginable piece of information (from a light beam carrying color to an indie rock concert to a single tree falling in a forest) can be expressed in the frequency domain exactly; the inverse transform can take a frequency description and exactly express it in the time domain (from negative infinite time to positive infinite time).
So why bother? What good is a Fourier Transform? There's no way to touch, see or perceive the frequency domain; it's an intangible concept. If you were to walk onto a college campus and argue about the tangible existence of the frequency domain, students would try to buy drugs from you and professors would laugh in your face - that's not a useful question because it's not something that can be known in our universe. The utility of a Fourier Transform lies in approximation - it can be "boiled down" from infinity to a usefully small amount of information and time, enough that the approximations can solve some quite interesting problems. Alternating Current (AC) electrical systems, guitar effects pedals/tuners, photoshop and missile guidance systems all make use of Fourier Analysis. Autotune (the reason all pop music sounds the same now) was originally developed as a Fourier Analysis program for oil mining.
By analogy, debating about the tangible existence of a soul (or the tangible existence of consciousness) is an endlessly circular debate because it doesn't have an answer. Our world of time, space, atoms and chemistry doesn't have room for a soul separate from a body, for consciousness without a brain. However, just because something is intangible doesn't mean it can't exist; just that the question of existence is poorly phrased (because it doesn't specify which universe you mean). In an analogy to Fourier transforms converting from our world of time to an imaginary realm of pure frequencies, I imagine a conversion from our world of atoms to an imaginary realm of consciousness (similar to Plato's realm of ideals). The idea has some appeal to it - every algorithm, every virus, every atom has some representation in this "thought domain". Every possible consciousness - every person who ever was, will be or might have been all exists in an intangible, invisible universe with no concept of time or space. When you die in the realm of atoms, your body may cease to function - but in the intangible thought domain your consciousness lives, along with every possible variation of your consciousness, your past consciousness from 5 years ago, the consciousness of just your amygdala on January 24th, 2005, your grandmother's consciousness before her stroke, the consciousness of a bacteria living on your left pinky toe, the consciousness of an app on your phone; everything. If you grew up religious you might recognize this description as a rephrasing of some very old concepts - to use some more math analogies, the total summation of everything in the thought domain into a single unified whole is sometimes referred to as "God". Rephrased, the concept of "God" involves the entirety of time, space, our universe and everything in it all functioning as a single unified consciousness of which we are all parts. Neat idea, impossible to prove right or wrong without resorting to circular logic.
So what good is this "thought domain"? How is it helpful? Also by analogy with the Fourier Transform - the important questions don't invoke existence, they are focused on utility. Colloquial usage for "does a creature have a soul?" is really better phrased "is it ok for me to abuse this creature?". Asking whether a particular collection of atoms deserves empathy and compassion - whether it qualifies as "having a soul" or "not having a soul" - is like asking whether a collection of frequencies qualifies as music. The answer is entirely subjective; there can be no objective definition of what music means (except Nickelback. I think we can all agree Nickelback doesn't count as music.)
Is it ok to kill a virus? Yes.
Is it acceptable to kill a flea? Yeah, that's fine.
How about a mouse? (Actually don't answer that, I might lose some friends on that question)
Historically, the evaluation of who "gets" a soul and who doesn't (and is therefore acceptable to abuse) was typically tied into political and economic factors. My prediction is that AI will only "get" a soul when it reaches sufficient complexity to fight for classification as such (think Black Mirror-esque brain digitizations). Like a Turing test, this is a messy and subjective way to do things, but I can't see a viable alternative.
So, as with the Fourier Transform, the really important question when thinking about consciousness and souls isn't "Do they exist?", it's "How is this definition useful?". If you believe in souls, religion, God, all that stuff - how does it make you act? Do you use that information to forge bonds with those unlike yourself; to seek commonalities and heal divisions wherever possible?
I've seen firsthand the power of religion to forge bonds, unite communities and strengthen those in need of help. Pretty much every religion focuses on Peace and Unity (the Arabic greeting "Salaam Alaikum" literally translates to "Peace Be Unto You" - the same phrase used extensively in Catholic services). However, every religion is a human institution; each has lots of biases and historical bullshit built into it (if that statement offends you, you can be assured that I'm not talking about YOUR religion. Yours is absolutely true and perfect. I'm talking about someone else's.)
Atheists can argue at length about the (non)existence of intangibles; again, I don't find the existence question meaningful. The more important question is utility - what does your atheism cause you to do? Does it mean that you spend time helping those who historically were crushed under the wheels of organized religion? Does it motivate you to show people that secular humanism can be just as much a powerful force of good as an organized religion? Alternatively, does it cause you to sink into apathy, isolation and nihilism?
Regardless of what beliefs you hold - do they make the world a better place? Do they make you a better person or a worse person? If your belief in Peace is exploited in order to justify hatred, then is it actually a deeply-held conviction or just words you say in order to fit into a community?
Comments
Post a Comment