Posts

Deepening the History of AI

The research project ‘Assisted Thinking’ pursues a long history of artificial intelligence by examining the interaction between scholars and their devices which help to organize their knowledge. This history spans a period of at least 350 years, gradually providing the building blocks that have led to the way that LLMs of the present function. The basic assumption of such a genealogical development is by no means to reconstruct the precursors of today’s technology of generative language models in a kind of teleological historiography, guided by the assumption that all these developments would have necessarily always culminated in LLM technology. Rather, the aim is to show how, under different technological conditions in past epochs, the relationships of users to their intellectual furniture have in each case led to new arrangements, practices, and descriptions that are characterized by the fact that jointly, through the interaction of humans and machines and through the distribution of their agency, they create an artificial intelligence in its own right.