Who does the work of keeping AI systems safe? This line of research examines the human labor behind AI red-teaming and evaluation.
Jan 1, 2024