Are People Computers? Strong AI, The Simulation Argument and Naive Realism
It has become very popular for people to believe that human beings are just computers. The idea that human beings are computers has several philosophical strands such as 'naive realism', 'strong AI' and 'the simulation argument'.
In the 'naive realist' approach all of our experience is considered to be the world itself. According to this approach, when you touch a cup the experience is thought to be a direct experience, it is thought to be the actual electrostatic forces between the physical finger and the cup.
The motivation for naive realism is usually the homunculus argument. According to this argument, if experience is not direct then experience must be an experience of data in the senses so we must perceive the senses rather than things in themselves; but how are the senses perceived? If the senses are perceived then something must be transferred into the brain to be perceived; but how is this data in the brain perceived? If this 'endless chain of relations' continues then there must be a little man with some magical property in the brain (a homunculus) that does the perceiving. Magic little men in the brain are impossible so the key relation that we call 'perceiving' or 'experience' must be external. According to Naive Realism experience must be embedded in objects themselves (see Note 3).
One of the most famous Naive (or Direct) Realists was Thomas Reid (1710-1796) although Foucault and many post modernists also believe that apart from objects themselves there is only an 'endless chain of relations' (Foucault 1966).
Naive realism leads to an idea of people as processors within a God-given array of things. There is no experience within the processors because 'conscious experience' is the world itself. This idea allows people to be characterised as computers because conscious experience is not a property of the processor; it is a property of the world.
Lay people who are naive realists are often unaware of the way that the senses operate and also of the physical nature of the world. The belief that what we sense is things in themselves usually originates in a misunderstanding about optics. The things that reflect light are mostly empty space; they have a tiny volume of electrons within them that create a field that reflects the light. The solar photons that are not absorbed are reflected from things in all directions and each eye selects and redirects these photons so that they form an image on the retina. If you hold a sheet of white paper up in front of the eye it remains white; there is no image on the front of the cornea. The only image is on the retina. What we see is based on retinal images, not things in themselves, in fact we can see silhouettes of things that reflect no light at all themselves because they form patterns on the retina. Furthermore we have two retinal images that can be quite different. FMRI studies show that these are combined in the brain. So what we experience is a neural combination of the input from both of the retinas; it is a pattern of brain activity based on data about things in the world that is called the 'perceptual field'.
We also experience dreams, hallucinations, imagination etc. as well as the perceptual field. These experiences are internally generated and share some of the areas of the brain that are used by the data that creates the perceptual field. Our experience is also a combination of touch, smell, sound etc. that are bound together as a geometric whole. Smells come from cheeses, sounds come from cars etc., this binding also occurs in the brain.
If data flows from the world to the retinas, ears etc. and from the retinas, ears etc. to the brain then our conscious experience must be at one end or the other of this chain of relations. The fact that the first pattern of things that is like our perception occurs in the brain and that internal events such as dreams, thoughts, hallucinations, illusions etc occur suggests that our conscious experience is at the brain end of the chain. So naive realists are correct in the sense that conscious experience cannot lie in the 'chain of relations', 'chain of cause and effect' or information flow but are wrong in assigning the phenomenon that is conscious experience to the disparate beginnings of the chain in objects themselves. Conscious experience is a phenomenon in the brain.
'Strong AI' is usually a form of Naive Realism although some proponents are Dualists (cf: 'emergentism'). The term was coined by Searle (1980):
"according to strong AI, the computer is not merely a tool in the study of the mind; rather, the appropriately programmed computer really is a mind, in the sense that computers given the right programs can be literally said to understand and have other cognitve states."
(Searle opposes the idea of strong AI). The basic problem with strong AI, apart from inherent naive realism or dualism, is that it relies on information systems theory. According to information systems theory it is possible to encode the state of a physical system in the state of another system. As an example, the state of the retina can be encoded in the state of a stream of impulses flowing out of the retina along the optic nerve or the electric potential of a circuit can be encoded as the displacement of a needle on a voltmeter. The carriers of the encoded state can be any physical things. The state of the physical things is called 'information'. A voltage, for instance, could be encoded as a set of dust particles at certain locations on a white sheet of paper. So long as the observer of the state knows where to look they can read the voltage even if the sheet were covered in other dust particles. According to strong AI the pattern of bits of information in itself is sufficient to create a mind. At any instant a system containing information, such as a sheet of dust particles could be conscious. Obviously such a belief depends on naive realism and would be laughable if it were not so widely held.
The simulation argument' is a mixture of dualism and strong AI. According to this argument (Bostrom 2003) the universe could be a giant computer simulation that contains people as well as objects. Bostrom seems to believe that at any instant a collection of bits of information like specks of dust on a sheet could be conscious, he states that:
" A common assumption in the philosophy of mind is that of substrate-independence. The idea is that mental states can supervene on any of a broad class of physical substrates. Provided a system implements the right sort of computational structures and processes, it can be associated with conscious experiences."
He then goes on to argue that we could be no more than simulations in a computer. Unfortunately, without tackling the problem of how a pattern of dust at an instant could be 'conscious experience' the simulation argument is flawed. In fact even a moving pattern of dust over several instants is problematical without the assumptions of naive realism or dualism. Bostrom, in evoking supervenience is probably a dualist; he puts 'mental' states' beyond physical explanation (ie: simply assumes that mental states could exist in a pattern of dust or steel balls etc.). This assumption is outside of the information systems proposed in the 'simulation argument' so the argument reduces to the proposal that the world is a simulation apart from something else required for knowing the world. This is not a proper argument at all - it is like claiming that the world is the simulation in my desktop computer if the entire world other than this simulation is excluded.
The belief that people are computers has not led to any scientific predictions and fails to explain the experience of even the most basic mental phenomena such as illusions and dreams. It seems to stem from either a desire to reject religion or, paradoxically, a religious desire to emphasise the 'oneness' of people with nature by asserting that our perceptions are the world itself. Adherents of this creed often state that 'there are no images in the brain' without realising that the images in the perceptual field originate in the CNS, starting with the retinas and extending to electrical patterns in the visual cortex etc. In fact, in normal perception there are no images in the world beyond the CNS!
Strong AI is problematic for several reasons. Firstly it holds that a set of 'bits' can be conscious. 'Bits' can be any identifiable thing, steel balls in boxes can be 'bits' so strong AI holds that an array of balls in boxes can have conscious experience (see Note 2). Secondly, from the outset strong AI assumes that conscious experience is a process, however when you look around the room conscious experience is a set of things laid out in space at an instant; conscious experience is a geometrical form or state, not a process. It is the acquisition of the data within conscious experience that is a process, not the experience itself. A geometrical state such as conscious experience is lost if it is encoded as a stream of bits in an information system, the stream of bits will no longer have the geometrical form that is conscious experience. Lastly, strong AI contains data in a form that only has meaning to someone who has a lookup table from the symbols in the system to their significance as a state (Searle's Chinese Room Argument - see Bibliographic Introduction and Note 1). It should be stressed that these arguments against strong AI/the simulation argument are not arguments against a physical basis for consciousness, a machine that replicated all the essential features of the brain could well become conscious but this consciousness would not itself be due to information processing. The contents of consciousness would be delivered by information processing but conscious experience itself is a state, a geometrical form. The machine would require a separate, consciousness module, that worked on different principles from an information processor.
The most pressing problem for philosophy, as well as science, is the nature of the conscious observer. What is the physical phenomenon in the brain that occurs at the end of the chain of relations, at the end of the processes performed by information systems? Unfortunately, despite our love of toys such as computers, there are no answers to this question in technological tools such as information systems theory and we must look to science for the solution.
See 'The Science and Philosophy of Consciousness' for an investigation of the phenomenon of consciousness.
Note 1: The "halting problem" is sometimes presented as a reason why humans could not be computers. The halting problem is similar to Searle's Chinese Room Argument in that transformations of information in an information processor contain states that are arbitrary, it needs an observer with 'conscious experience' to determine the significance of a set of transformations (such as the infinite loop of the 'halting problem'). To put it another way, the 'halting problem' is a rewording of the problem of 'the endless chain of relations' or the 'homunculus problem', without a particular state at the end of a chain of relations any attempt at simulating observation will result in recursion. The issue for scientists is to determine the nature of the 'state'; it is not a simple 3D array of things.
Note 2: The difference between information as an encoding of a state and a particular state itself is geometrical. If an encoded state has an identical geometry to a particular state (congruent at every level) then it is a replication rather than a simulation. This suggests that the 'Knowledge Argument' is actually a form of the 'Simulation Argument' because the colour red studied by Mary is only identical to the 'red' in perception if it replicates the form of this 'red' exactly - 'red' encoded in a non-congruent form is a different entity.
Note 3: The scientific conclusion would be that the theory that information flow creates experience is wrong, information flow must merely provide the data for experience. In science theory must be consistent with observation.
Links
Thomas Reid. Stanford Internet Encyclopedia.
http://plato.stanford.edu/entries/reid/
The Simulation Argument by Nick Bostrom
http://www.simulation-argument.com/
Bibliography
Reid, Thomas. 1764. An Inquiry into the Human Mind on the Principles of Common Sense. Pennsylvania State University Press, University Park 1997.
Searle, J.R. 1980. Minds Brains and Programs. The Behavioral and Brain Sciences, vol. 3. Copyright 1980 Cambridge University Press.
http://members.aol.com/NeoNoetics/MindsBrainsPrograms.html
Bostrom, N. 2003. Are you living in a Computer Simulation? Philosophical Quarterly, 2003, Vol. 53, No. 211, pp. 243-255.