JakeH comments on Is AI safety doomed in the long term?