intrust

Most W3 works better as dW3. You might prefer an original oil painting to its digital copy, – no matter how fine the resolution of the copy. You might prefer experiencing a symphony orchestra live as opposed to hearing it on a compact disc. I, and most people I know would much rather read a book in its traditional form rather than scrolling through its pages on a computer screen. But if we wish to do things with W3 other than first hand personal consumption, if we wish to copy, to analyse, to edit, modify, compact, index, store, compare, calculate, encrypt or distribute – then dW3 works best.

Since we are already familiar with a metaphor for layers we might as well reuse it again in the context of dW3.

The bits – the ones and zeros of binary logic are the track all dW3 must travel on. The representational syntaxes; the alphabets of written languages, the spatial-temporal encoding of waves of light, electricity and atmospheric pressure, these are the flatcars. The containers are the grammatical rules of each media. And the goods? The goods are the meaning, the message, the expression – the raison d’être for all W3. It is possible to compute at any of these four layers, but the complexity involved increases as one moves upwards in the stack.

Computers can do a great deal with digital bits without having the slightest idea of what they represent. They can dumbly store them and dumbly send them places, dumbly discover redundancies in their sequential patterns and compress them, or dumbly encrypt them to make them secret. This is the railroad track level.

At the flatcar level, the computer must know something about the encoding syntax of the various media forms of dW3. The computer must know that A,B and C are letters of an alphabet, and that certain combinations of bits represent colours or fonts or the amplitude of a wave of sound. At this level computers can sort and filter, index, and hyperlink. More clever forms of data compression and encryption can be employed.

At the container level, computers can spell check and grammar check, flesh out the beats-per-minute in a musical recording, make assumptions about pictures¹ and find even more sophisticated methods for compression and encoding for purposes of secrecy.

And then there is the goods level where the computer can carry out tasks only possible though some understanding of what it is working with. To compute at the goods level the computer must be able to derive the meaning and intent. In other words computers must be able to think. But can they?

// posted by Greg FitzPatrick @ 14:10