Turing machines, one kind of stuff and artificial consciousness.
Computers seem to be so good at so many things.
They are able to calculate with accuracy and
efficiency
that very few humans could ever hope to match.
They foster communication and connection in a manner which even some of the most
complex social structures find difficult to attain. This has set me wondering
whether there will ever come a time when computers are able to outperform humans
in that third type of knowledge, spiritual intelligence.
In their superb book
SQ
Danah Zohar and Ian Marshall suggest that there are three types of intelligence.
Firstly, there is IQ (Intellectual Quotient). This is the kind of intelligence
that has to do with logic and reason. It applies certain rules in a very linear
way to come to particular conclusions. In fact they suggest that this kind of
intelligence operates within the human brain through a series of neural
connections (neurons are the cells in the brain that fire the electric charges,
or currents, that make the brain work) that are connected in a linear fashion.
People who posses a high degree of this kind of intelligence can do calculations
quickly, the operate well with rules, and are able to make fairly concrete,
black and white, decisions. Of course computers can do this very well. They
operate according to preset (or preprogrammed) rules e.g. if this happens then
do that, if that happens then do this, if neither happens then do this or that
(this process has become known as the Turing process, after it's designer Alan
Turing). Because of the fact that this kind of intelligence works well with
rules it did not take too long for computers to be programmed that could do
things, which required an ability to operate within the constraints of certain
rules, very well. For example the chess playing super-computer developed by
IBM, Big Blue, which beat Gary Kasparov, the world champion chess master at
Chess. Give a computer enough accurate programming and it will be able to
adequately figure out what response makes the most sense. Add to that the
processing power to perform these calculations with great speed and you have a
machine that will outperform a human, within the ambit of it's programming,
every time! However, change just one variable and the computer will be stumped.
You see it can only operate within the limited confines of the program that has
been fed into it.
The second kind of
knowledge that Marshall and Zohar identify is EQ (Emotional Quotient). Of
course, anybody who has read recent works in corporate culture and personal
development should be familiar with this kind of the knowledge. This is a kind
of knowledge that allows one flexibility to make creative and diverse choices
within the confines of certain preset rules and conditions. Whereas linear, IQ,
only allows one to make the choices of the program, EQ allows one to make
choices within the scope of the program. So, if this isn't so, and that isn't
so, it doesn't mean that I crash and stop working. Rather it means that I find
another answer that works in order to solve the problem. This kind of knowledge
is not linear, it is parallel. Within the brain it is suggested that humans
have the capacity for this kind of knowledge because of extremely complex
parallel neural connections. For example, I am learning to ride a bicycle and
fall off. A logical thought process would say "you cannot ride a bicycle so you
should not climb on one, since because you cannot ride, you will fall each time
you climb on the bicycle". However, a complex, or parallel, thought would ask
"are there any other instances that I can draw on, from other experiences that I
have had, where I have learned how to do something that I could not do before?
Yes there are, here is how I learnt these skills and abilities, so apply
something similar from another context to this context [e.g. trial and error,
perseverance, drawing on the knowledge and skill of others etc.] and I will
learn how to do the thing that I cannot do know." Thus, even though it is
logical and reasonable that I may fall off the bike again, because I cannot
ride, my mind tells me that through processes that I already have some
experience of I can learn how to ride. So I climb on the bike again. Computers
are able to do this task well. They have two things in their favor. Firstly,
they are able to store, or remember, things well. If information is stored it
remains useable as long as it is made available to use, and of course as long as
nothing goes wrong with the machine and wipes out all the data (a.k.a use a Mac,
not Windows!) Neural networks and Artificially Intelligent (AI) machines that
are programmed with the ability to alter their own code, or programming, in
response to certain circumstances, are examples of this. For example, some
companies use AI machines such as those mentioned perviously to manage trades on
the stock market. A machine may be programmed to automatically sell all stocks,
or buy on more stock, if the stock level reaches a certain level. However, the
machine also stores 'experiences' of the outcomes of previous trades. For
example, the machine may store that three out of four times when the stock price
suddenly dropped below a level when it had been programmed to sell off all
stocks, it suddenly rebounded to a much higher level than it held before the
fall. Thus, because of this 'memory', the machine alters it's programming to
say something along the lines of "don't immediately sell when the stock reaches
this level, first wait a day to see whether it rebounds, if it does not then
sell, if it does start to rebound then buy". You can see that it is issuing
itself and instruction which may be contrary to the initial instruction that the
human programmer has given. But, at the end of the day the computer's new, or
changed, instruction makes more fiscal sense. Where as a human trader may panic
or grow impatient and make the wrong decision, an AI machine should become more
and more accurate in it's decisions to sell or buy, the more experience and data
it has to store. Again, add to this process increased speed and you have a very
accurate, highly efficient, machine that could outstrip a human being in
EQ.
The third kind of knowledge that
Zohar and Marshall speak about is an integrative knowledge, which they call SQ
(Spiritual Quotient). This is a knowledge that works not only with the rules
(like IQ does), and not only within the rules (like EQ does), it works the rules
themselves! Let's use another hypothetical, and very simplistic, example. A
person lives in an oppressive society. If such a person only had IQ, they would
either have to obey, or not obey the rules of the country. If the person had EQ
as well, they would have to try and find ways of living within the rules
(finding exceptions and flaws in the rules which to exploit). However, a person
with high SQ would seek to live outside of the rules, maybe even creating a new
set of rules. IQ asks, "How can I do it?" EQ asks, "What can I do with it?"
SQ asks, "Is this what I want?"
Now
clearly, this kind of 'transcendent' knowledge is not yet a capability of the
computers that I use (although, I must confess that my MAC does seem to defy
many rules!) However, the question that one needs to ask is why is it not
possible, and just because it is not possible today, does it mean that it is
impossible? Think about it, just 150 years ago it was not possible to phone
another person, to fly, to drive a car, and a myriad of other things which are
commonplace today.
One of the strongest
set of arguments that are given for why this kind of intelligence will not be
possible for machines are arguments which are based upon variations of the
understanding of human consciousness. Many argue that machines will not be able
to do this kind of thinking since they are not conscious. They are not
creative, they are created. Many argue that the reason why we can apply SQ
within our lives is because we are conscious beings, we can think, but more
importantly we can think about ourselves. In other words, I can ask myself,
"how do I feel about this, can I do anything about it, do I have to live this
way..." More importantly I have something which is known as 'metacogition', the
ability to think about my thoughts. A computer can only 'think' this, or
'think' that. It may even be able to alter it's 'thoughts' in some way (as
mentioned above). However it cannot think whether the thoughts themselves are
valid or not valid.
This is where the
theorists fall into two camps. In his book
God and the
mindmachine John Puddefoot speaks of the
monists and the dualists. Let's first talk about the dualists. These are
people who say that mind and matter are two different and distinct things. Like
the philosopher, Descartes, they say that mind is something separate and
distinct from the physical works (res
cogitans versus
res
extensia). Plato, of course, was one of the
earliest recorded thinkers along this line. He believed that people were souls
that were trapped in physical bodies. Within the Christian tradition we have
many such neo-Platonic ideas (particularly those of the Gnostics). I have also
noticed that forms of neo-Gnosticism are prevalent in many modern Charismatic
Churches that emphasis the importance of the spirit over, and against, 'the
flesh' (which is regarded as weak and sinfull). Anyway, the dualist argument
holds that machines, which are matter, could never become truly conscious since
they are a completely different 'stuff' to mind. They are physical and not
spiritual.
The other camp are known as
the monists. They are people who believe that everything is one and the same
'stuff' (see some of the papers that I have written on this website at
http://www.spirituality.org.za which refers to such thinkers as the Quantum
Physicist, David Bohm, and the monk Dom Bede Griffiths). There is a fare amount
of current scientific theory that suggests that mind and matter come from the
same common spiritual source. Of course as Christians we should hold to such a
view if we take texts such as Ephesians 1:10 and Collosians 1:16-17 seriously.
In short, the proponents of this view, whether Christian or from other faith
traditions (particularly faith traditions that are not dualistic - such as
Hinduism and Buddhism) hold to the notion that since all reality is of the same
'stuff' there is no reason why consciousness is not possible for something that
is material. After all, we as human beings are matter and we are conscious.
Thus, some theorists have applied variations of this view to suggest that
machines (whether electronic, mechanical or biological) have the same capacity
for consciousness, and thus transcendent or spiritual existence, as we do.
Sure, this is a very strong view of Artificial Intelligence, it borders on manic
optimism, but it is logical if one agrees with the underlying principles and
thought processes.
So, it could be
possible that machines could one day be better than human beings in all three
spheres of intelligence, IQ, EQ and most importantly SQ. Popular films such
'The Matrix', 'Dark City' and of course the Stanley Kubrick classic '2001 a
space odyssey' have all speculated to the outcome of such an eventuality. If
evolution continues to operate, even at the level of consciousness, then it
could be possible that humanity would become the inferior species. Thus, at
worst we could face extinction, or at best be harnessed (ala The Matrix and Dark
City) by machines for some menial task to sustain their
life.
There is of course a far more
optimistic approach to this possibility. This approach is based, in large, upon
a variation of Pierre Teilhard de Chardin's notion of evolutionary cosmology (I
have written something on that which you can find on my website at
http://www.spirituality.org.za). de Chardin suggested that the whole of the
cosmos is in a process of complexification, this evolutionary move can be traced
from the dawn of time right into the future. In essence he postulated that the
cosmos is evolving from the less complex, gross or material levels of reality
(biogenesis) to the more complex and subtle spiritual levels of reality
(noogenesis) to a point of ultimate consciousness which he called the Christ
Omega or Christ consciousness.
Within
such a model it is possible to assume that any move in complexification is a
positive evolutionary move. Thus, any increase in consciousness is of benefit
to the cosmos as a whole. Of course this view is non dualist, in that it
supposes that everything (persons and the rest of creation) are all of one and
the same stuff. Hence, it may be plausible, if one supports a notion such as
this, to suggest that even if humanity does become extinct, or the lesser
species in creation, this may be part of the evolutionary plan of the cosmos as
it moves to a higher plane of
consciousness.
I'm not so sure about
all this. However, the one thing that we cannot deny is that the boundary
between technology and human persons is quickly diminishing. Not only are we
becoming more dependent upon technology for our very survival (and here I am
thinking both of life saving technologies such a biomedical, mechanical
lifesaving devices such as pacemakers etc. and simple technologies like computer
that control currencies, electricity, and other day to day
functions).
The one other question that
is prominent in my thoughts is the question of when the crossover takes place
between machine and person. The movie 'Bicentennial man' clearly illustrates
the difficulty of judging this from the machine side. It asks the question,
because a machine looks like, behaves like and has emotions like a human person
does that make it human? The question is also asked very pointedly by Ray
Kruzweil in his book The Age of
spiritual machines, when he asks how far must
a human person go before he or she is classified as a machine. For example, a
person who has a Cochlear implant to assist them to hear would be regarded as
human. A person with artificial limbs would be regarded as human, even a person
with an artificial heart and vital organs is regarded as human. However, if we
were able to take this technology to its extreme (which of course is not yet
possible, but could at some time be a possibility) and do something along the
lines of what some theorists suggest could become possible, i.e. downloading our
brain and all it's thoughts, ideas, memories, feelings etc. into a computer,
would that mean that the computer becomes human? Or does it mean that even
though 'I' may still be the conscious element of the machine, because I am not
biological, or largely so, that I am no longer human? What then if one uses a
computer that is biologically based, using enzymes to process the code of 1' and
0's, rather than a silicon based machine? Does this make a difference?
These are some of the thoughts that
occupy my mind in the wee small hours of the morning. I suppose I won't mind
too much, as long as I don't become a Windows box!! ;-)
It;s all been said above....
Posted: Thu - April 22, 2004 at 05:44 PM