David Harket comments on Too Many Metaphors: A Case for Plain Talk in AI Safety