Friday, May 20, 2016

leavebots on the line - how to tell when a brexit campaigner isn't a "real" person

so if you engage with online campaigns, you will be aware that much of cyberspace (the twitter sphere etc) is full of bots - it is also full of people who are paid, or lined up one way or another, to "manage" the discussion, rather than let it flow from the population at large.

At one level, this is good, as it can help prevent trolling or other forms of abuse.
At another level, it is bad, as it puts a (deliberate) chilling hand around the throat of human discourse - indeed, methods were developed by Putin in Russia and by the government in China to police social media, and these are showing up around the US and UK online political debating channels increasingly. Amazingly, the linked study of chinese control of media gives 5 examples of methods, and here I paraphrase very slightly, so you can see how much these are used in the brexit campaign social media content:

1. Taunting of Foreign Countries, to take critique of this one:
e.g.  This country "has ijk million people, the more they rise, the more difficult it will be for us to eat, because the earth’s resources are limited."

2. Argumentative praise or criticism
e.g.  "The system automatically had you follow xxx & yyy populist online social media users,
this is a standard tactic of indoctrination and brainwashing,
I suggest you unfollow"

3. Non-argumentative Praise or Suggestions, that are a bit off topic - e.g. (yes, in china)
"...the housing problem"

4. "Factual" Reporting
e.g.  "we suggest you follow the opinion of leader C, with his selfless dedication to moral character and hardwork" - remind you of anyone (boris?)

5. Cheerleading
e.g. "Many heros fought bravely to create the life we have today - respect these heros" - used by both sides relentlessly

A simple clue is to look for "automatic" use of trigger terms - for example, in the brexit campaign, there are some tropes setup by the campaign managers (accuse critics of being a fear or scare monger, of being deluded, or of just being on the gravy train, etc etc

If a response doesn't address anything specific in your content, then it is suspect. If it is long, but appears so soon after your comment on fb or twitter, then likely it was cut and paste from a campaign document - check for this, as then you can draw attention to any of these behaviours, and detract from them, or indeed, get people to start filtering content coming from such non genuine participants.

Once you've learned the ropes, then you may get some rewarding discussions from thoughtful people, and maybe change your mind, or theirs. And that's how participatory democracy works, folks......

At some point, I might fire up our botnet detection software on some of these sites, but I am reluctant to start reporting them for abuse, as every now and then it looks like a real person has just quoted from stuff they've seen unwittingly, but that doesn't invalidate that person's view one bit. well, only just a bit...

No comments: