A method of generating subjective time in machines
February 7, 2011 1 Comment
For computers, there’s a system variable that is used to compute the time. this moves in an incremental and regular electronic way. Humans, however, have a non-standardized perception of time (notwithstanding the cultural and social forces that mediate our conceptualization and perception of time). This may be an important germ of self-awareness for living things. Just as the game of ‘peek-a-boo’ serves to educate infants that 1) the precipitants of sight remain relatively constant when the eyes are closed and 2) others have lines of sight distinct from ours, the experience of sleep, a sudden rush of adrenaline, or the passage of years can all contribute to a conscious being’s understanding of his/her temporal perspective.
To parse this perspectival nature, we need an element upon which the rest of the system is paced. In humans, the experience of rhythm performs this function (here I conceive of rhythm as the persistence of a stimulus within short term memory such that an impression of it remains and can be contrasted with present stimuli). A non-sentient computer uses the system variable to pace itself. Leaving aside the possibity of distributed computation/hybrid computational storage systems, we can still alter a computer’s system time in a non electronic way and mimic temporal subjectivity. If the system time were 1) correlated with the measurement of a rhythmic variable such as a heartbeat, daylight intensity, reletive humidity, etc 2) varied according to a complex mathematical function of Wolfram Type 4 automaton or 3) some combination thereof, we could start to experiment with its effects.
The subjectivity of time cannot be perceived by the subject, however, unless it is contrasted with an objective rhythm. We therefore need to bring the subjectively ‘entranced’ machines above into contact with other similar machines or systems and then, through a process of entrainment, override the rhythmic, subjective time, and synchronize them, a more or less jarring experience.
My sense of the experience of de- and re- synchronization is based on my own periods of getting lost deep in thought and then returning to another ‘speed of existence’ when interrupted by someone else. I’ve also had the experience of mutual entrainment where I realized later that time passed much more quickly than I had thought.
Some practical considerations for this proposal include how to handle faults and errors generated in the linear processing scheme by changing variables of time. Again I feel led towards distributed processing (e.g. neural networks) for a successful realization of this idea.