A Grand Idea

Watching Steve Grand lecturing yesterday, or more specifically watching his answering questions at the end, was like being a spectator at a fencing match, as he deftly parried and stabbed at challenges about the nature of consciousness with his sharp intellect. The creator of the computer game Creatures, once said to be the cleverest man in Britain, was arguing that what we know as consciousness is something of a sham, a highly predictive simulation of the way our world is at the moment we experience it. This "multiple drafts" model is Daniel Dennett's idea in Consciousness Explained, and I have a lot of time for it, as it explains many of the blips in our perceptions. Anyone who has ever tried, with frustration, to draw a coin will know how effectively our mind informs us that a coin is round and thus it is the roundness we try to transfer to paper, rather than the oval we actually see. Grand used the example of your brain tracking an eagle dropping from the sky; because it takes 500 milliseconds for the light from the eagle to reach your eye and pass through the interpretive receptors in your brain, by the time this processing is complete the eagle will not be where it was when you first "saw" it. The world we see, and the world we know, are not one and the same.

The problem I had with Steve Grand's ideas, however, was his argument that if you modelled an electron on a spreadsheet (inputting data for co-ordinates in space, charge, etc.), then modelled enough electrons in combination to form a hydrogen atom, then enough to form a two hydrogen atoms and an oxygen atom, then enough to form many trillions of these, then you would have "made" water vapour. This, following on only from the fundamental rules you described at the molecular level would, without no further intervention from a deus ex machina (i.e. The Programmer) eventually condense and fall in the form of rain which, eventually, would form a lake, rivers and so on, all behaving almost precisely as rivers do in the real world. It would be so precise, in fact, that the resultant entity, because it follows precisely the same laws of physics at the real thing, simply residing in the software rather than reality, would be a lake. However, the challenge I would make to this is how would I know when I was looking at the digital lake that it was a lake, and not my garden pond, or the Red Sea. In order to know the lake was a lake, I would need to put it in a context that defines "lakiness," namely it is probably large, not artificial, and often exists in the basin between or at the end of river valleys. In order to produce a digital lake that was real enough to be called a lake one would also need to model the environment, including soil, hills, light, sound that permit our senses to interpret it as such.

With a similar level of recursion, how would one define when conscious life has been created artificially? To model this, one would need to take the serial processor of the computer, which would model a parallel processor, which would model evolving life, which would evolve to model consciousness which is, as Grand says, only a simulation anyway. Where, on this long scale of replication, is the oval office where the buck of being stops? It would be easy if we had a firm knowledge of where life - and especially conscious life - begins. It would make ethical issues such as determining whether to withhold the life support of a severely handicapped baby that much more regular and straightforward. Sadly, though, the problem is not, nor will ever be, that easy. Whether it will be cybernetics and artificial intelligence, or discursive philosophies of mind that will come closest to a definition of consciousness, watch this space (it may take a while...).

Labels: consciousness, Science and Culture, Steve Grand

0 Comments:

<< Home