Tag
A preprint proposing Verification Learning Theory (VLT), a formal process model of how AI-assisted claim checking affects retention of independent judgment, predicting that perceived assistant fallibility can improve delayed unaided performance through evidence checking.