I have come to the very strange conclusion that I have nearly spatially isolated in myself an adjustable algorithm to create word-selection attitude (maximum orientation setting, from the set of vectors pointing inward from the bottom of a vertically oriented circle, being the diameter that points straight up; I call it Matt Malcheno). Of unfortunate note is the fact that there is in fact one and only one totally viable way to inflect any particular word-selection attitudinal mode, but of fortunate note is that there is a subset of word-selection attitudinal modes with a cardinality of 4 that adequately charters any precise flow of words from any particular desired underlying meaning. This means that one still can play within a (possibly shifting) inflection subspace=word selection subspace, while giving a speech, for instance, once one knows exactly what words one wishes to use. If I could break this algorithm down to American English, if not more rigorous, code, I could probably subsequently garner the assistance of someone else to write something that might pass a notable subset of textual Turing tests. The algorithm to adjust the orientation vector is of course perhaps equally tough, if not tougher, to write, because it must evaluate the (possibly shifting) range of orientation vector possibility of another writer, and then be responsive in a way that seems human.
Posted on: Mon, 05 Jan 2015 19:27:27 +0000