| HN Mirror

You only have to write the argument once. Then use the social media management software that's being peddled to tweak the wording slightly so you have prepared text stating the same thing 50 different ways. Then if someone gives a counter-argument that you have a prepared response for, you can copy-paste that in a matter of seconds. Arguments you haven't prepared for are ignored because the goal isn't to engage in dialog, it's to give the false impression that a dialog is occurring. To an outside observer the forum would appear to be populated equally by people for and against the topic. Even though it's really an overwhelming support for one side and a single agent spamming with 50 different personalities.

Derailing a conversation works well because they will have a large number of responses prepared that talk about, using your example, Snowden being a traitor. So once they've wedged the issue their spam can become the dominant voice. If the discussion had drifted to an area they weren't prepared for, say the historical precedents for whistleblowers of government misdeeds, they probably don't have as much material for that.

My point is I assume astroturfed material would not be written off-the-cuff but meticulously edited ahead of time to give the desired impression. And near the top of that list must be the requirement that it not look like an obvious shill.

I wonder how effective it would be if agents were able to control both sides of the debate, even? A more sophisticated shill would give the appearance of supporting the opposition but will subtly help draw attention to the propaganda. A living straw man, as it were. Not unlike the way SWAT police will pretend to be violent protesters to goad troublemakers into doing something they can be arrested for.