It seems to me that line of argument is refuted by what the experiment is; he made a post, then looked at the replies, and their timing. Twelve hours of no vie

It seems to me that line of argument is refuted by what the experiment is; he made a post, then looked at the replies, and their timing. Twelve hours of no views on Twitter, then a mass of hostile robot posts, then his post is generally visible and starts getting views. BlueSky, immediate responses and all in favour of Ukraine. This isn't the algorithm choosing what he views. These are what other people posted in reply to his post.