[JumpCrisscross]

> This is true for every subfield I have been working on for the past 10 years

Hasn’t data labelling being the bulk of the work been true for every research endeavour since forever?