Rhyss comments on A response to Matthews on AI Risk