A mind is like a parachute
It might save your life,
but you have to know how to use it first.

Friday, June 29, 2012

You can't argue with a sock puppet

This is going to come up later, so I want to get it down right now.

This Wired Article suggests that the military is quite keen to get in on the social media scene.  Not just to detect when "rumors" get started in battle zones (or maybe an uncomfortable truth or two leaks out) but also how to respond to these "memes" in real time.  Now I don't like the word "meme" and I will talk at length at some point about why I think merely "signal" or perhaps "social signal" is a better word, but "meme" is the going concept these days, and since this is just a concept mini-mart, I have to stock what sells.  [For anyone who doesn't know a meme is "an idea behavior or style that spreads from person to person within a culture".]

So the goal is to be able to see what's out there in the social media space and get out there and respond to it (or spin it) as necessary.  This can include the creation of "sock puppets" to spread the word.  A "sock puppet" as the name implies is a fake account which is controlled by another person (or piece of software) which acts as a mouthpiece for someone else.  A computerized "sock puppet" response system would use a bunch of artificial accounts (which appear to be genuine) to "shout down" any unwanted viewpoint or message.

To give a crude example, let's say that a village school was hit with a drone instead of the intended target (for whatever reason whether bad intel or targeting error or what have you).  Many kids died. If there is a social media footprint in such an area, the reaction could be swift and severe.  The local net could blow up with all kinds of outrage.  "8 children dead. 20 wounded" could be one tweet, with a link to photos of the damage.  "Does this girl look like a terrorist?" could be another, with a picture of a sweet looking (and very dead) four year old.

News travels fast on the web.  But if there were a way to respond, "robo-spin" could automatically step in and produce just enough ambiguity to prevent this from being too big an issue.

Jabavut (sock puppet 1): I was near the school and I saw 3 men leaving the bldg just before the explosion.
PeaceMoon7 (sock puppet 2): Really?  Did you see what they looked like?
Jabavut (sock puppet 1): They were dressed as police officers, but my guess is they were Taliban.
Shpun11 (sock puppet 3 -- Shpun is the Pashto word for "Shepherd" lending an air of authenticity to the fake identity):   They have done this many times.  Disguises to Police
SunWarrior750 (sock puppet 4):  We need to find out exactly what happened.  This is awful.
HopeMission16 (sock puppet 5): Jabavut, I just missed you today,  I am so glad you are safe!

This conversation could be created instantly and plugged into the emerging social media response.  It provides a false account to produce ambiguity, an inquiry that validates the original observation and provides greater "detail" to the false account.  A "local commentary" is added, complete with broken english.  Then a demand for "the Truth" prompts the casual reader to reflect on the many possible things which might have taken place and gives them all a sudden equivalency they would not otherwise have, and finally, the last sock puppet merely feigns interest in the well being of another sock puppet, leaving the casual observer to believe that these two people knew one another and that there was at least one other person who could vouch for the legitimacy of Jabavut's "observation".

Scared yet?

Well the issue gets more complex than my example would lead you to believe.  Suppose the "enemy" was engaged in precisely the same thing, and the school had in fact been blown up by locals who wanted to blame an American drone attack?  How would the truth compete with the social media attack designed to spread misinformation very quickly?  The sock puppet is like any other weapon -- it can be used to defend as well as attack.

And any way, as the Wired article points out:

Darpa’s announcement talks about using SMISC [in] “the environment in which [the military] operates” and where it “conducts operations.” That strongly implies it’s intended for use in sensing and messaging to foreign social media. It better, lest it run afoul of the law. The Smith-Mundt Act makes pointing propaganda campaigns at domestic audiences illegal.

Phew, that's a relief.  Because without that law from 1948, domestic audiences (we the people) could be subject to the same kind of misinformation propaganda that the Pentagon might use in a battle zone.  Only, it appears that through a quiet little amendment in the defense authorization bill, that protection against propaganda may be about to die.

Now is the time to think about being scared.

No comments:

Post a Comment