AI : genealogies of
Méthode critique de Kate Crawford : “excavating AI” (title, 257), “research an archive” (181). Annotations, collection, labels. She says “genealogies” somewhere – concerning how data training sets are reused, hypotexts, by different projects (and in contest settings, curiously, a from in this tech history) and in cases different generations of.
Also why, or how, history of science and history of tech are so relevant. They too, predating in their epistemology the new objects of contemporary tech, and thereby avoiding the warp of presentism. (They chart that warp itself in fact, it’s formation.)
History of media also placed here, but with the care of “remediating” “media” – à la John Guillory, in a classic philological, literary-theory methodology of tracing the successive synonyms and developing semantic & designative fields. The jump from one plane to another, and the actors of the shift, sudden expansion, inclusion, englobement. McLuhan major logothete here.
Another key : reintroduce the disciplines. Against their erasure under de historicism get tech-talk.
And of course, in the institutional context of US AI development : reintroduce the role of universities – and the structural set-up of security & science, military-industrial-academic complex, of the 40s, 50s and beyond.