Norbert Wiener’s God and Golem, Inc. is a discus­sion that almost touches upon the confronta­tion between science and religion. The reason it doesn’t quite manage this is because it is restricted to the field of cyber­netics — the science of commu­ni­ca­tion and control — and the resulting exposi­tion deals with the relation between man and machine, which is no longer as relevant as it was around fifty years ago.

Wiener’s argument is an aggre­ga­tion of ideas that deal with three major aspects of machines -

• Machines that learn
• Machines that reproduce
• Co-ordination of man and machine

Learning, says Wiener, is something that makes sense only when applied to a problem that cannot be completely under­stood in a reason­able amount of time. Take a game of chess, for instance. If combi­na­to­rial explo­sion did not prevent us from predicting every possible set of moves, we would be uncon­quer­able as opponents, and there would be no scope for learning from past games. If the omnipo­tent God were to play a game of chess with a mere man, what chance would the mortal have? None, because every move of his would have been precal­cu­lated and the result would have been inevitable. What then is the purpose of the game? It is nothing but a charade, and for the purposes of discus­sion, it is of little use. It is there­fore neces­sary to avoid the notion of infini­ties such as omniscience.

Machines learn. There are computers that play chess, and do a lot many other things efficiently. These machines can lose a game once, yet avoid doing so again; this is the process of feedback associ­ated with learning. But, quite contrary to this spirit, he questions the idea that machines can be given exclu­sive decision-making authority, arguing that humans are the sole judges of what is right, and no amount of learning can fully under­stand this. True, a machine remains a machine, and morality is defined by its associ­a­tion with the human factor, but to shy away from speaking of man and machine in the same breath is unnec­es­sary in this era. We live in a world of technology, and technology is synony­mous with machines; the protest of religion against mecha­niza­tion is not fatal for the pioneer.

The process of repro­duc­tion in machines is curious, because Wiener sees it as dupli­ca­tion with varia­tion. Not only is the dupli­ca­tion a neces­sity, but so is the varia­tion, although there may be varia­tions within the varia­tions. Indeed, this is phylo­ge­netic or racial learning, if we assume that varia­tions improve our opera­tional abili­ties. Again, we have the ethical question before us — does man dare play the role of God? God made man in His image, and he made the machine in his, thereby displaying his arrogance and defiance towards the Divine. I don’t believe this view is supported by many other cultures — Hinduism speaks of God as the eternal Self, immutable, formless and blissful — so I don’t think it lends itself to deeper analysis. And even if it did, ethics is an issue that can only be judged by its times; it evolves the way language does — wrong usages are accepted once the novelty wears off.

Finally, there is the issue of co-ordination or inter­facing between man and machine. Modern technology can replace parts of the body with mechan­ical devices that serve the purpose, and technology of the future will be capable of doing more. Do what you best, says Wiener, and let the machine what it can do best. Wiener prefers to consider humans superior to machines, appar­ently, although there is a reason for not doing so. Superi­ority, or the lack thereof, is neces­sarily a human perspec­tive rather than an absolute notion, so opinion is a factor that counts. But from a purely Darwinian point-of-view, the survival of the fittest can be under­stood as the fight for survival of the individual rather than a race. If the help of machines will ensure victory, then that is the route to be taken. If machines learn to defeat the humans and take their place as the dominant race, then that is what we call evolu­tion, because learning is indeed the crux of the evolu­tionary process.

In summary, Wiener sees a world built upon a large number of simple rules, a problem that can be solved with appro­priate program­ming, but I doubt if things could be that easy. I see the rules of the game as something akin to Mandelbrot’s fractals, non-linear, inter­de­pen­dent, ever-changing and haunt­ingly myste­rious in its intricacy.

This entry was posted on Thursday, March 23rd, 2006 at 10:21 and is filed under Daily Rant. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.