Generalist, Outlier AI — AI training and response evaluation
You rated AI answers by comparing which response was better and explaining why. You also fixed or rewrote responses to improve quality and correctness. You checked outputs for mistakes, bias, or nonsensical content to support safer and more reliable AI behavior. • Evaluated multiple AI responses for relative quality • Performed error, bias, and coherence review • Rewrote or corrected responses when issues were found • Authored prompts to test and probe AI behavior