Aside from the different perception between the genders. It goes without saying that these biases must be accounted for even before we start training AIs on these classifications.

do or don't? Your Mastodon post says "don't", your Bluesky post says do.

DESI has the cosmology survey, other science piggybacks on.
This one defines surveys first and then distributes the observations to the best targets available from a selection function (finish targets which need various visits, cadence requirements, survey completion percentage, etc.).

I forgot to say that the 15 surveys are on top of the 10 Consortium surveys, so 25 surveys in total.

Targets are:
Metal poor halo stars, Stars in the Magellanic Clouds, eRosita / AGN follow-up, Time-Domain targets, Cosmology, etc. etc. etc.!