CommentStreams:48a3af61d354ead38e7fe8e9b7344fc6

Thanks for this contribution, Sami, is really interesting and I was quite clueless about IAA and their relations with models like LAION.

Perhaps what I'm missing to understand this paper is the context where IAA happens. Is there an (explicit) relation between the museums and IAA? or does IAA mostly happens parallel to museums or similar cultural institutions?

Pasquinelly's work on e.g. neural networks and models (*The Nooscope manifested: AI as instrument of knowledge extractivism*) emphasises that dimensionality reduction is an inherent part for creation of models. In this case, the creation of an AII score is not derivative but basic, since a model "standardise" as an internal technique. What I found particularly interesting is what is arguably gained by this: is the "aesthetic" of images standardised in a more distributed way (by e.g. distributed tagging), and thus, less "reduced"? (I hope this make sense!)

Like Paul, I'd probably suggest a more emphatic conclusion, otherwise, the main point is a bit lost in a (very interesting) discussion.

(I wonder about the table, and if you want to have it in the proceedings, but we can talk about this when we meet. Oh, also, are the "two wolves"" are a reference to Deleuze/Guatarri, and if so, why not make it explicit? (or perhaps it's just a coincidence!)