What do we mean by reproduction after all? At least in sexual reproduction it does not imply the production of exact copies of the original(s). It is precisely the variation of / the aberration from the original that made sexual reproduction successful (quality vs. quantity) and less vulnerable to infection, more dynamic than simple cloning or apomixis (the Red Queen hypothesis; Darwin's early warning against monocultures and affirmation of variation). So, maybe we should mean by reproducible science, the R in FAIR, not exact replicability, but rather the possibility of varying experiments or test conditions, and still receiving similar results. Remains to be defined the level of similarity, then. [Interesting to note also the somewhat different distinction between reproduction and replication made in the Neurips2019 Task description, where replication is defined rather as reproduction from scratch, without using the same code as the original, what would in reality seem to be closer to a variegated, "sexual" form of reproduction, on the face of it.]
Kommentarer