142 Environments of intelligence
What can be claimed for the first subclass of cognitive artefacts but not for
the second is that the domains of the signals produced by the artefact are convergent with the domains of natural information that is or could be available to the
receiver in a given situation. The term “convergence” assumes a technical meaning in this context:
(DS-1) Convergence is achieved if, for each domain D a of a type of signal r a
that is emitted by some artefact and observable for R, there is a domain
D n of another type of, ultimately natural, signal r n that can be, or hypothetically could be, observed by R while both domains are supposed to
be correlated with the same conditions F at the source s, and transformations thereof, by the same or corresponding regularities of informational
mapping.
If information basically is the co-occurrence between a proximal and a distal
world affair in which conditions at the distal end covary with conditions at the
proximal end in accordance with a strict natural regularity (see NI-1, Chapter 4),
and if some artefact is supposed to track the same conditions at the distal end
that the unaided receiver seeks to keep track of, although possibly by different
means, the condition will still hold that any transformation of conditions at the
distal end shall be mapped by transformations of conditions at the proximal end.
If the artefact, in order to track the distal condition, records the values of the same
proximate variable as the human receiver by means of a sensor that is built differently from the respective human sense organ, the mapping regularities will be
virtually identical for D a and D n , even if that sensor is more receptive or capable
of recording a wider range of values than the human sense organ. If, however, the
artefact uses a different proximal correlate of the distal affair in order to track the
same conditions at the source that the human sense organ is in pursuit of, transformations in the respective proximate correlates will have to correspond in order
to meet the condition of convergence. If conditions at the source change, either
proximate variable will have to change in such a fashion as to not only ensure a
mapping onto distal conditions individually but also provide a reliable or, for the
purposes at hand, reliable-enough mapping between the transformations of the
values of the natural and artefactual proximate variables.
For example if the pattern of visible radiation of some distal object were used
as a cue of a certain condition of that object by the human observer while patterns
of thermal radiation of that object were tracked by an artefact, the values of the
variables recorded by the human observer and those recorded by the artefact are
bound to correspond only for some objects. Although a source of electromagnetic
radiation emitting a spectrum of waves quite similar to that of the sun, as filtered
through the Earth’s atmosphere, will preserve correspondence in this respect, anything emitting a different spectrum of electromagnetic waves may not. LED lights
will be found to shine very bright without getting very warm. A piece of iron will
get considerably hot in the process of being welded or forged well before emitting
a visible glow. Hence, either the range of objects to be tracked has to be restricted
Précédent

- 155/231

Suivant