More on Representation [reader]

AI is essential to perform computation over scientific datasets, yet
representation is defined upstream (one step before) of AI itself

The article below [1] provides a good background and brings up important
questions on
the notion of representation at large in science (the material  that AI KR
works with)
In the same context,
Also, essentially reinstating the purpose of this CG, is [2]
Read, discuss

*[1] Representation in scientific practice revisited 2014*
https://www.academia.edu/8505481/Representation_in_scientific_practice_revisited_2014
As the first volume to bear the name Representation in Scientific Practice
(Lynch and Woolgar 1990; hereafter, RiSP)
demonstrated, representation involves lengthy struggles with research
materials to
reconstruct them in a way that facilitates analysis, for example through
coding and
highlighting key features of interest and aligning them with particular
concepts and
theories. This treatment of representation in and as practice has since
spurred a rich body of ethnographic, historical, and discourse-analytic
inquiries that demonstrate how the circumstances of knowledge production
are folded into epistemological claims and
ontological orderings

[2] *Representation in AI Evaluations*
https://dl.acm.org/doi/fullHtml/10.1145/3593013.3594019