Linda Linsefors comments on Interviews on Improving the AI Safety Pipeline