We're training AI on our actual conduct which is... shading results, burying findings, making sure nobody notices.

https://www.ianbetteridge.com/the-worst-of-us/