Well, then you should have noticed what the article was about, which was not to detail a research program about AI safety. Different articles can address different aspects of a problem without being accused of advocating "random action". That's just ridiculous.