When two digital people meet for a chat in Second Life ®, there are at least six people involved in the conversation.
To help explain the reasoning behind that statement, I shall introduce a hypothetical digital person, known as ‘Digi’. Digi needs a friend, and so we need another digital person. Here comes one now. She is called ‘Tal’.
And then there were two.
In the essay ‘Virals And Definitives’, I explained that ‘some… see SL as most strongly linked with novels, theatres and movies… technologies that can organise patterns of information in such a way as to make you or I believe in the existence of somebody or something that does not necessarily exist in RL’. Just as every literary character owes its existence to an author, and every puppet requires a puppeteer, so behind a digital person we consider there to be someone out there somewhere in RL, typing away at a keyboard. Digi is no different. He knows there is a puppeteer working Tal, and Tal knows there is an author crafting Digi’s words and actions. Carrie Fisher once noted, ‘I am not famous; Princess Leia is’, drawing a line between her life and that of the iconic character she once performed. Carrie Fisher and Princess Leia are two different people. Something like this seperation exists between a digital person and its primary.
Two have become four, but we are still missing two other people.
But think about what has been happening in your mind as you read these words. You can hear a voice in your head. It is the voice of your inner self, the voice that subjectively expresses your thoughts. Only, the words you hear being spoken are not YOUR words. The thoughts those words convey are not YOUR thoughts. They are MY words expressing MY thoughts. I have used written language’s extraordinary capacity to trigger language circuits of the brain and in doing so I (in some sense) have become you, or, you have become me.
The same thing happens whenever Digi and Tal converse via text in SL. As Digi reads Tal’s words, language circuits in the primary’s brain are activated and, even if only partially, Digi is Tal (or Tal is Digi). Moreover, human brains are machines complex enough to have achieved stage 4 of Stephen M. Omohundro’s 5 stages of technology:
Stage 1: Inert systems, defined as anything that is not actively responsive to the environment.
Stage 2: Reactive systems, respond in different but rigid ways in the service of a goal.
Stage 3: Adaptive systems, change their responses according to fixed mechanisms.
Stage 4: Deliberative systems, can construct internal models of reality and choose actions by envisioning the consequences.
Stage 5: Self-Improving systems, can comprehensively redesign itself and is able to deliberate about the effects of self-modification.
In SL, the vast majority of objects are inert systems. After all, just about anybody can rez a few prims and combine them to produce an object that sits there and does nothing. (It might serve a useful purpose. A shoe is useful, but it is still an inert system). Reactive systems require a modicum of scripting talent. You might design a box that people step into that can act as an elevator to take them up and down to different levels of your store. Some companies like Daden Ltd have created automated avatars and other narrow AI applications that qualify as stage 3 technologies.
As for stage 4, the most advanced robots in the world are beginning to cross this threshhold, but we still have no technology that can match the human brain in terms of general intelligence. Digi’s primary’s brain is a machine capable of building internal models of the world, including models of people he has met (obviously, it can also model fantasy worlds and characters, or how else would fictional works exist?) Whenever Digi is in dialogue with Tal, he uses his internal model of Tal in order to determine the next best course of action.
Actually, it is only this internal model of Tal that Digi knows. The brain, after all, is not in direct contact with reality and everything that people normally consider to be ‘real’ is actually a simulation created by the mind, based on information gathered by the senses. Natural selection would have surely selected against incorrect models of reality, so it is safe to assume that much of what we believe to be real is indeed a very close approximation to actual reality. But it is also well-known that a person’s beliefs about Life, the Universe, and Everything are not always in agreement with other people’s, as any study of theoretical physics, or philosophy, or theology, will show.
Reality generates overwhelmingly more information than any one brain can hope to store and process. What an individual comes to believe in is shaped largely by the information they happen to have been exposed to, and the way they have been brought up to interpret that information. ‘Truth’ may be defined as ‘Whatever is actually the case’, but when someone declares something to be true, its is much more likely that they mean ‘this is compatible with my prejudices’.
Digi’s model of Tal is affected by the information that already resides in the primary’s mind. He cannot create a model of Tal that perfectly matches the model of Tal that is stored in Tal’s primary’s mind, because the two of them have not shared the same life experiences. Their past experiences were not so different that they cannot relate to each other, obviously, or else they would not be friends. But, strictly speaking, the Tal that Digi has come to know exists nowhere but in his mind.
The brain of Digi’s primary therefore runs three models, three patterns of information that conscious awareness perceives as minds. One pattern becomes known as ‘I’, another as ‘Digi’, and the third as ‘Tal’. Where in the brain do concepts of self and the intentions of others, form? How are they formed in the first place? How, in other words, do we explain the mechanisms of the mind?
Steven Pinker began his book, ‘How The Mind Works’, by flatly stating ‘we don’t know how the mind works’ (the book goes on to argue that our understanding of the brain has progressed from a mystery, ‘not knowing what an explanation would even look like’, to a problem, ‘we may not have its solution, but we have insight… an inkling of what we are looking for’). Since we don’t know how the mind works, we might as well entertain several different states of mind for Digi…
THE SERIAL MIND
In this scenario, only one theory of mind is running at any given moment. So, whenever the ‘Tal’ model is active, the other models are either completely inactive, or have slipped so far into the subconsciousness that they effectively do not exist. It is easy to see how an imaginary character like Digi or Tal ceases to be as soon as the mind focuses its attention on modelling other things, but the idea that ‘I’ ceases to exist is somewhat harder to accept.
Then again, one can think of instances where people seem quite detached from reality. Patterns of light projected onto a screen, patterns of ink embedded on a page, result in us bursting out in laughter, or wiping away tears of empathy for characters that do not exist. In ‘Multiplicity: The New Science Of Personality’, Rita Carter explained that ‘David Suchet had a long stage run in ‘Timon Of Athens’ during which he found it increasingly difficult to flip back (to being David). One evening, a psychiatrist friend… shot at him a number of questions such as: ages of your children? Phone number? Date of birth?… The ‘Timon’ personality was so firmly in charge that his… memories were temporarily irretrievable’.
Of course, not everyone earns their living or spends their days roleplaying historical or fictional characters. But, as Rita Carter argued, ‘when someone says “I don’t know what got into me”, or “I just wasn’t myself today”, they are implicitly acknowledging the existence of a self other than the one who is speaking’. Perhaps, in normal circumstances, the brain can switch so rapidly between theories of mind that ‘I’, ‘Digi’ and ‘Tal’ seem to co-exist, when in reality if ’Digi’ is running the others are offline. Alternatively…
THE PARALLEL MIND
In this scenario, the three models are running simultaneously. On one hand, this scenario makes more sense than the previous one, since it denies the possibility that ‘I’ ceases to exist when the ‘Digi’ or ‘Tal’ model is active in the brain. Also, any lecturer explaining the computational theory of mind is sure to point out the massive parallel processing capabilities of the brain, which goes far beyond those multi-core chips that are all the rage these days.
On the other hand, what we have here is a conflict with most people’s basic concept of ‘self’. Rita Carter called this concept ‘Oneness’, as in the belief there is ‘a single, consistent, identity’. In stark contrast to this belief, the Parallel Mind scenario posits three mental models of identity running simultaneously: The primary is ‘I’ and ‘Digi’ and ‘Tal’. Of course, one can always claim that two out of the three do not exist. But they do exist as patterns of information interpreted into a model of a person, just as ‘I’ does. And once you start down that road you wonder if sounds do not really exist, unless and until pressure waves rippling through air molecules are funnelled into ears which then trigger auditory circuits, or if streams of photons are not light until they strike the retina and the visual cortex builds up an image of a scene. And if you were someone like Morgaine Dinova, you would insist that sound waves and photons do not exist either, and are just convenient scientific models of some ultimately unknowable reality.
THE NESTED MIND
The Nested Mind scenario sees the model of ‘Tal’ existing within the model of ‘Digi’ which in turn exists as a model within the mind of ‘I’ which in turn exists… well we will get to that. Before that, I should point out the difference between ‘Parallel Mind’ and ‘Nested Mind’. In the previous scenario, the primary’s model of ’I’ is equal in size (abstractly speaking) to the models of ‘Digi’ and ‘Tal’. According to Nested Mind, ‘I’ is biggest, ‘Digi’ is smaller and ‘Tal’ is smaller still. ‘Digi’ exists within the model of ‘I’ because, well, it was the primary’s life experience that built up Digi’s i-genome (see ‘Virals And Definitives’ for the definition of i-genome). And ‘Tal’ is nested within Digi because it is only through Digi that the Primary knows Tal. So you might also call this scenario ‘Matrioska Mind’ after those Russian dolls that split open to reveal progressively smaller dolls inside.
The question of where the mind exists has traditionally met with two conflicting answers. There is dualism, which professor Owen Flanagan also calls ‘Nearby’, because, ‘when I try to remember…it seems as if my mind asks my brain… (‘I’ live where my mind lives, not where my brain is housed)… the mind is not in the brain or the body, but it is close’. The idea that a person’s mind floats around in some metaphysical space close by, working the body like an invisible puppet master deftly plucking invisible strings, is a weird but popular one. It is, after all, the default position of anyone who believes the Self survives the death of the body and brain. Most people do believe in an afterlife of some kind or other.
No such comfort is offered by the other answer, which professor Andy Clark calls ‘Brainbound’. According to Professor Flanagan, brainbound insists, ‘the mind IS the brain. The mind isn’t in the brain in the metaphorical way that my beloved has a place in my heart. The mind is literally, physically in the brain because, well, that’s where it is and that’s where it was all along’.
Andy Clark himself has introduced a third alternative that he calls ‘Extended’. According to this view, ‘Mind’ is a pattern that is smeared across many brains and external memory systems. In the introduction to Clark’s book (which is called, ’Supersizing The Mind’), David Chalmers writes, ‘when parts of the environment are coupled with brains in the right way, they become parts of the mind’. Suppose you save records of important appointments on your personal organizer, and come to rely on the machine to recall information you entered and subsequently forgot (why retain so much information, when you only need remember which few icons to click in order to get the machine to ‘remember’ for you?). Well, in that case the personal organizer counts as part of the cognitive process that reliably results in you keeping your appointments. No wonder an interviewee once told Sherry Turkle, ‘when my laptop crashed, I felt like I had lost a part of my mind’.
One can think of all kinds of examples that seem to agree with ‘Extended’. Your facial expressions, body language, and the meaning conveyed by your words trigger sympathetic responses in my brain. I model your mind, you model mine, I know that you know that I know how you feel, so both our minds are smeared across more than one brain. A pianist reads the score for Beethoven’s ‘Moonlight Sonata’, and her brain produces the same patterns of motor-neuron activity that makes her fingers dance over the keys in the same way that the great composer’s fingers once danced. When she performs to an audience, patterns on the page trigger patterns of brain activity that trigger patterns of finger movements that cause the musical instrument to trigger patterns of jostling air molecules that are translated by the minds of the audience into beautiful, melancholic music that, if only slightly, lets them know something of how Beathoven was once feeling.
Andy Clark’s hypothesis also raises the possibility of a meta-mind existing within and distributed across several information-processing systems. The character ‘Digi’ is not the result of one person’s mind, because other residents in the online world interact with, and therefore influence, the development of that personae. According to the Extended perspective, if Digi and Tal exist anywhere, it is in the abstract space that exists between the screen and the minds of the people who connect via the Internet to online worlds. As Philip Rosedale once speculated, ‘Second Life… could be looked at as one collective dream’.
If Second Life is artificial in the sense that it was imagined and constructed by collective human activity, the same must be true of our RL cities and countryside. The only difference is scale. SL emerged from the collective imagination of a few million people over a period of years. Cities, from the emergent activity of tens of millions over a period of centuries. The countryside, from the agricultural activities of hundreds of millions over millenia. Apply the Nested model to SL as a whole, and we discover it is a network that exists within the larger network that is the Internet. The Internet, in turn, arose from external memory devices known as cities. Essentially, society and culture are gigantic memory storage systems, a vast extended phenotype, a super-organism that emerged from the collective activity of the human race and its technologies.
Among many other things, the super-organism has introspected on questions regarding the nature of mind. We now believe that the brain should not be thought of as a single organ, but an intricate network of many regions that have specialised pattern-recognition capabilities. According to Ray Kurzweil, work in cognitive sciences, robotics, miniaturisation and biotechnology may one day enable brain prosthetics to evolve into a network of nanobots that can read patterns of brain activity, supress neurons from firing and cause neurons to fire. If my sense of self arises from the patterns rippling across the network of regions in my brain, what would happen if, say, we took several hundred people living in California with nanobot brain prosthetics, have region 1 in person 1’s brain copy activity in region 1 of my brain, person 2 copy region 2 and so on? (note that only one region in any one person’s brain is copying the patterns of activity in the corresponding region of my own brain). Suppose that each region can, thanks to prosthetic abilities, communicate with the regions in other people’s brains. If ‘I’ emerges from the coordinated activity of several hundred regions in my brain, would the same coordinated activity distributed across the brains of several hundred people mean ‘I’ am a self-awareness hovering over the State of California?
Imagine that advanced nanotechnology has resulted in dust-sized supercomputers embedded in all physical objects, and the age of ubiquitous computing realised by networking these computers together with sensors for gathering all conceivable information, is upon us. The Web has become co-existent with the world. It is everywhere, immersion achieved not merely through headsets or plugs in the back of your head, but by the now omnipresent nature of the Web, a true Digital Gaia. A child is born and his whole life experience is continually uploaded to the network where a model of the child’s body and brain is located (asking where, exactly, it is located makes no sense, since it exists in no one physical place but rather is smeared across the dust of ubiquitous computing.) The infant moves an arm, and the process of building a corresponding body map of that limb in the mind of the infant, is begun (let us assume nanobots are so ubiquitous, even babes in the womb do not evade their omnipresence). Would a perception of a phantom limb, and a phantom head, and a phantom heart, also emerge within Digital Gaia? If the child grew into a teenager who died, would downloading the emulation of that person into a convenient android shell be a realisation that the corporeal self survives physical death?
Really, these are different variations of old thought experiments. From Robert Nozik’s ‘Experience Machine’ that provides virtual but compelling experiences of one’s heart’s desires, to Descarte’s ‘Evil Deceiver’ that tricks us into believing in a world that does not exist. From ‘My Grandfather’s Axe’ (father replaced the handle and I replaced the head, but it is still regarded as ‘Grandfather’s axe’) to the ‘Ship Of Thesus’ (over the ages, it is repaired by replacing old planks but is regarded as the same ship even when not one original part remains), questions regarding identity and the nature of mind and reality have piqued our curiosity.
Roleplaying in online worlds, artificial intelligence research and cognitive neuroscience are some of the modern technologies that have been pressed into the service of answering the questions, ‘How does the mind work’ and, ‘where is the self located’? To say we are on the verge of finally answering such questions may be arrogant, given the heavyweight scientists and philosophers who have, in the past, dedicated their lives to answering these questions only to fail. But we can say that questions that have occupied our minds for millenia, stored within the external memory systems that emerged from society and culture, will not be deleted from those memory systems. Moreover, the network of humans and machines are becoming increasingly interconnected, increasingly capable of information processing, and, just maybe, edging ever closer to some kind of enlightenment.