"Epistemological Lessons from the Automation of Science" Prof Alexander Bird (University of Bristol)

Science is increasingly automated. Automatic weather stations and satellites have for some time collected raw data which is supplied directly to computers for analysis, whereupon weather maps are published on the web while the analysed results are also fed into meteorological and climate models. DNA sequencing, once a lengthy and expensive process involving considerable human input, is now almost entirely automated, where automation includes both the bio-chemical intervention with a sample and also the statistical analysis of the results of the biochemical assay. In this paper I focus on two sets of questions: 1. How should we understand `observation' in automated science? I argue for a functional rather than aetiological notion of observation. 2. What is scientific knowledge? I argue for a social conception of knowledge, where the `social' includes scientific infrastructure as well as scientists.