So. Many. Fucking. Bots. (And they’re good)

Ugh. This can be considered a follow-up to my post yesterday on the same issue.

On Facebook today, I saw the following link in my “trending” section.

Screenshot 2018-03-26 16.25.54

The story is essentially just this one picture that’s been circulating around Twitter:

Well, isn’t this some great liberal content! Red meat for the base. Well — knowing what I know about deception on Twitter, my first instinct these days is to be skeptical about the integrity of the information I’m getting. And, sure enough, if you go to @LazyyMillennial‘s twitter timeline, it’s been made very recently and devoid of any real personal content.

Very much not surprisingly, if you run this account through the same botrnot model I used yesterday, you get a very high likelihood (~92%) that this account is a bot.

Screenshot 2018-03-26 17.30.33

“The Hill” is a pretty well-known outlet, and as of the time I checked out the page 2,937 shares. And, of course, they are not the only ones that picked up on it.

HuffPo is carrying the story, as well as the Houston Chronicle. All of these outlets are credulously referring to this Twitter account as “Rebecca” even though they have not gotten any kind of quote from her (they only quote what’s posted on her profile). Some are even featuring a screenshot of her photo in a kind of Ken Burns moving image with narration video.

Now, perhaps the picture is real (Huffpo’s story had the image tweeted out by someone else), but “Rebecca” is certainly helping to amplify the message:

Screenshot 2018-03-26 17.32.13

Is this a bot in the sense of a purely automated account? I don’t think so. It seems to me like a human writing these quotes. But is this person the person they claim to be? Seems very, very unlikely.

Once you start looking for bots — or impersonated accounts, which may be a better term, they are everywhere. That last piece of content you chuckled at while wondering why {conservatives, liberals} were so dumb? (“Yeaaahh we gottem good!”) Or that made you really angry, and made you wonder how “the other side” could think that way? That may well have been manufactured. That may not be real. There seem to be a lot of “Rebeccas” out there.

 

Our discourse is still being poisoned

Today on Twitter I came across a tweet that seemed so heartless, so inhumane, that I had to wonder whether or not the perpetrator, was, in fact… human

For those who don’t know, Patrick Petty is a father of one of the Parkland victims. Unlike most of the people that have emerged and become quite visible out of Parkland, Patrick is conservative and pro Second Amendment.

So, the tweet above is beyond inappropriate. But if you look further, something seems a bit off. If you check out @simonsaysboohoo‘s profile, it’s a mélange of emotionally-charged liberal content. It’s… exclusively a mélange of emotionally-charged liberal content. There’s nothing else.

Using the botrnot model developed by Mike Kearney, this account has a very high probability of being a bot (~92%):

Screenshot 2018-03-25 20.23.34

The big problem of course is that a lot of people are going to think that @simonsaysboohoo is a real person, reflective of an “other side” where people actually do think these objectionable and immoral things. And why would they think otherwise? A person is not privy to the summary statistics of user behavior that the model above is using to calculate the probability of bot-hood. We are enormously easy to fool.

So, someone is trying to inject poison into our public debate; but who? and why?

And probably most importantly, why isn’t Twitter (and the same applies to Facebook and Google) doing more to prevent this kind of deception?