Thursday 12 April 2012

Dream of a sleeping machine

My previous post on silence is making me think about the possibility of a machine that 'appreciates' silence and inactivity. Most interventions in robotics are, after all, about activity, not inactivity. Yet it is often in inactivity that we find meaning (as for example in the 'inactivity' of performing 4'33''). Bearing in mind the description of the inactivity of experiencing silence that I presented before (http://dailyimprovisation.blogspot.co.uk/2012/04/what-is-meaning-of-silence-expectations.html), what would it take to create a machine that similarly would recognise the regularities in itself, attend to the regularities in others, and appreciate the random sounds that disrupt and transform those regularities?

This would be a machine that sleeps, snores, turns, coughs and gurgles. How could that be done?

My first thought here is that the mechanisms of the body are tied up with the mechanisms of the mind. Sleep is an attenuation of the senses and the sense-making processes. But whilst our eyes are shut and we don't react to environmental sounds, smells and sights in the way that we do when we are awake, our nerves carry information about our bodies all the time.

Some of this sensory information is regular and some not. Heartbeat (especially) and breathing (mostly) are regular. But digestion, bowel contractions, mucous  reflexes and body positions are not. (It may be that the change in body position relates to digestive, bowel or mucous reflexes). How could we look at this kind of sensory information?

This has made me think that after so long trying to create artificial brains in computers, we ought to turn our attention to creating artificial bowels. The fluid dynamics of digestive juices can be modelled. With that the information arising from a modelled bowel can be simulated.

But how might this help in the creation of a machine that sleeps?

The point of all this is the relationship between the meaning of silence and the meaning of sleep... and beyond that, the meaning of meaning. Looking at meaning in a sleeping machine, there would obviously be information transfers between the regular events and those which are irregular, much in the same way that the 'musical' experience of listening to John Cage shifts between regularity and disruption. The body responds in sympathy: or at least the conscious (or semi-conscious?) body responds. An event in the gut produces a disruption to breathing, a change in position... a flood of sensory information where before there was little of note. And then it subsides and regularity takes over again. There is information transfer (in the Shannon sense); there is a restructuring of expectations. There is meaning.

The sleeping machine is not an inactive brain. It is a brain wired into gut, heart, lungs, kidneys, liver, muscles, mucous and blood. The sleeping machine is a machine intent on its search for meaning among the sensations of sleep.

How is that different from a waking machine?

And if sleeping and waking are both searches for meaning, and meaning is the central objective, why oscillate between sleep and being awake?

My only initial thought to that question is that sleep is a disruption to being awake, and being awake is a disruption to sleep.  

No comments: