Linch comments on Talking publicly about AI risk