View Single Post
  #2 (permalink)  
Old 08-30-2007, 06:59 AM
David Bailey David Bailey is offline
Senior Member
 
Join Date: Aug 2007
Posts: 1,144
Default

I think this thread is an excellent idea. I would like to suggest that we keep it just for personal position statements, and that we don't just spark another version of the endless debate here If nothing else, this would dilute its content and make it hard to read.

I began life as a chemist in the UK, where I still live. However, I did so much computer programming as part of my PhD, that afterwards I decided to move into software development.

Over the years, I have developed a number of compilers (programs that translate computer source code into machine instructions), including, in particular, a compiler for the Artificial Intelligence (AI) language, PROLOG.

This was in the 1980's when AI was all the rage. I guess I accepted the commonly held view that the race to produce a truly intelligent computer program would soon be won. The idea that you could duplicate the 'software' of the brain inside a computer seemed amazing, but not impossible - just as many physics theories seem incredible at first sight (curved space-time, quantum theory, etc).

I think the general enthusiasm for AI at the time was fuelled (in me and others) by the following logic. The brain is just a biochemical/electrical machine - a sort of fault-tolerant computer - and as such, it could in principle be simulated on a computer - which would then constitute a continuation of the brain's consciousness. The SF writer Greg Egan is particularly fond of stories of this type in which people transfer their consciousness to computer hardware.

Of course, any actual simulation of the brain was no more than a gedanken experiment, but if such a thing were possible in principle, it seemed reasonable that a much simpler program could implement thinking as such without getting involved with the exact details of how nerve signals are passed across synapses, etc. There is an interesting analogy here. When Intel develop a new type of CPU, they write a program that simulates the operation of the CPU chip at a very low level. They actually load conventional programs into this simulator, and check that the simulation of the new CPU actually delivers the right answers!

Although that work on the compiler did not involve AI directly, I saw a lot of other people's programs and read a fair bit about AI. I gradually became completely disillusioned by the whole field. The very fact that the gold standard for an AI program was the Turing test - in which a candidate program would try to fool a human that it too was human - seemed to sum up much of the research - pretence and wishful thinking. (This is not a dig at Turing, who developed his idea in the very early years of computing.)

BTW, if anyone thinks the subject of AI has not failed - just read some of the hype of the early 80's (repeated in the 90's when people tried the neural net paradigm). It is also worth reading one or two of Greg Egan's SF stories - such as "Permutation City". I believe Greg Egan is a believer in strong AI, but parts of his stories read like gedanken experiments demonstrating the implausibility of the concept.

To me, this left a huge hole in the conventional scientific picture - usually referred to here as scientism. If the vast amounts of funding that went into AI could achieve so little, maybe there was something wrong with the basic logic of the project. Far better minds than mine have also seen that hole - Roger Penrose's two books "The Emperor's new mind" and "Shadows of the mind" are very well worth reading, and reflect the struggles of a top theoretical physicist to understand consciousness.

OK - that is the background behind my interest in PSI and consciousness. I feel I have watched scientism fail, and that makes me more open to alternatives. I always used to think of PSI as total bunk, but in recent years, I have been amazed at the hatred of conventional science towards the paranormal. There is very little money for PSI research - so relatively few experiments get done, those that do get done are smothered in statistical argument, and if that fails - as in the case of NDE's - extremely dubious explanations (brain activity while flat-lined) are touted almost as if they were established fact. Furthermore, as you can see if you visit Brian Josephson's site, the sceptics can even use totally underhand tactics to discredit PSI. Furthermore, science is incredibly reluctant to make the links between PSI and accepted science. Libet's neural timing experiments are extremely hard to explain, but are accepted as genuine. That puzzle is obviously related to the presentiment experiment - but nobody wants to know!

Reputable scientists that publish anything paranormal can suffer something akin to the inquisition. Jacques Benveniste was a respected biochemist until he published evidence that might indicate that homoeopathy was genuine. Without wanting to get into that debate, I want to point out that his career was destroyed by the witch hunt that Nature instigated in return for publishing his results! There is endless talk - here and elsewhere - about the concept that many negative PSI experiments are never published, and that this skews the statistics. My suspicion is the opposite - that many experiments that show a PSI effect are buried because the experimenters do not want to ruin their careers!

**** Afterthought ***

I forgot to mention that I was brought up a Christian, but lost my faith at university. I am now pretty much against the orthodox religions because I feel they have discarded any spiritual quality and become mere vehicles for (usually male) power and aggression. I don't see religions where you are taught what to believe as a useful way forward.

David

Last edited by David Bailey; 08-30-2007 at 12:22 PM..
Reply With Quote