Outlier AI projects
Occasionally I would get Data Science related/Mathematics projects where I would have to compare the outputs of different responses from two different models and compare them against a rubric, then I would have to write a justification as to why I chose that outcome/rating and back it up with evidence from the response and rubric. A few projects were related to programming languages and mainly just reading outputted code from a model and doing write ups on whether or not the code would function as intended (which was based on the prompt), if there were any errors, and finally, I'd have to go back and edit the output if I thought it was incorrect.