Is this Twitter Bot trying to affect the national zeitgeist, or just spam?

[Mick West]

The twitter account @jenn_m_tweets is a bot that randomly tweets every few minutes a simple opinion randomly constructed along the lines of:

"I think that (contentious subject) is (positive/negative) and (that's good/bad)

Some examples:

In my heart, I believe alex jones is bad
I think feminism is important
MRAs are shit
TERFs are straight terrible!!!!
I really believe that andrew auernheimer is uncool
richard dawkins is shit!!!✌️
IMO, safe spaces are important !!!!
In my opinion, trump is 'cray cray'

I first saw the bod because I've got a Tweetdeck filter set up for chemtrails, and I saw the tweet: "chemtrails are fake and that's important" pop up.

Source: https://twitter.com/jenn_m_tweets/status/997556305857515520


Which has been repeated for quite some time, along with thousands of tweets on other topics

The bot appears to tweet statements that are designed to anger people in what would loosely be described as alt-right. People who listen to Alex Jones and Milo Yiannopoulos.

So there's a couple of questions. Firstly why does Twitter allow it? Has it just slipped under the radar?

What is the bot for? Is it some social experiment? Part of some nefarious Russian active measure to increase dissent? Just spam to make you click on associated links? (it never posts links). It follows mostly German Tweeters, which might be a clue. But I really can't figure out what it might be for.

 

[deirdre]

it thinks:

no one seems to be liking or sharing it. maybe 'they' made like a thousand bots, hoping a few would get some followers and then just forgot about this one?

 

[marrowmonkey]

The bot appears to tweet statements that are designed to anger people in what would loosely be described as alt-right.

It seems like an alt-right caricature of a "liberal", I don't know much about US politics though. It seems a bit obvious it is a bot, although it is possible to get surprisingly good results with a simple set of rules. Maybe it doesn't have to be more sophisticated to be effective at whatever it is aiming to achieve?

So there's a couple of questions. Firstly why does Twitter allow it? Has it just slipped under the radar?

Bots are not against Twitter's rules.

The study, which used more than 1,000 features to identify bot accounts on the site across six categories, estimated that between 9 percent and 15 percent of Twitter's active monthly users are bots. Using Twitter's most recent figure of 319 million active monthly users, that translates into 28.7 million to 47.9 million bots.

https://www.cnet.com/news/new-study-says-almost-15-percent-of-twitter-accounts-are-bots/

It can be really difficult to detect bots. In MMORPGs there are always those who use bots to cheat and it's a constant battle between the developers of the game and the developers of the bots to try to outwit each other. Maybe twitter decided it was easier to just allow them and only ban the ones they think are causing a problem and get reported (i.e. spammers)?

 

[Svartbjørn]

The image also either appears to be an older or photoshopped image of Lotte Verbeek. I ran the profile pic through image search and thats the closest I could find. It may not make a HUGE difference in what you're looking for @Mick West but it might give you a reference point to start with more digging. Many of the russian hack bots Ive run into on Instagram etc, have tended to use either European models or old images of European Actresses to gain followers.

Spoiler

 

[Hevach]

What is the bot for? Is it some social experiment? Part of some nefarious Russian active measure to increase dissent? Just spam to make you click on associated links? (it never posts links). It follows mostly German Tweeters, which might be a clue. But I really can't figure out what it might be for.

IF it's part of the Russian bot network, they did play both sides of the table, just the results were more immediate and glorious on the far right far better than the far left.

The whole purpose seemed to be signal amplification of the extremes, probably hoping they would feed of each other in an outrage feedback loop. Possible end goal was simply shouting down every reasonable voice so that any political debate would inevitably become wing nut vs. moonbat. Because they did feed both ends, a bot meant to incite one side could fail at that but still succeed in outraging the other.

This one looks like it did neither, but bots are basically free - you can throw a bit of everything at the wall and only really need a few here or there to stick.

 

[derwoodii]

What is the bot for? Is it some social experiment? Part of some nefarious Russian active measure to increase dissent? Just spam to make you click on associated links? (it never posts links). It follows mostly German Tweeters, which might be a clue. But I really can't figure out what it might be for.

could it be just a test bot like a survey or a web farming exercise to gather respondent comments and accounts for later use.

Im always amazed at how many Bots there are out in the web and curious what they are up too?? See below from another site i frequent,, 61 robots..

1. derwoodii

Total: 303 (members: 21, guests: 221, robots: 61)

 

[Hevach]

In world wide web terms, a "robot" and a "bot" are different things. A robot is a program that reads websites, a bot is something that generates input as if it were a user. Robots include search engine webcrawlers, news aggregators, some kinds of feed application (I'm not sure if the one I used back in the 90's still exists or how widely they're still in use, these days I can just open 5 or 10 or 100 browser windows and get lost instead of needing to budget my page loads for dialup time), etc, so there's a lot of them out there. A site can be configured to tell the difference between a robot and a user so that they aren't generating spurious page hits or user counts.

 

[FatEarther]

The image also either appears to be an older or photoshopped image of Lotte Verbeek.

No offence, but that 'bot' pic looks nothing like Lotte Verbeek. If Mick used that as a reference point, he'd be wasting his time.

 

[Trailblazer]

No offence, but that 'bot' pic looks nothing like Lotte Verbeek. If Mick used that as a reference point, he'd be wasting his time.

Yes, I couldn't see the resemblance either!

It's actually a crop from an Alamy stock photo. You can even still see the watermarks on it.

http://c8.alamy.com/comp/C8EYKA/two...e-girls-one-blonde-one-brunette-uk-C8EYKA.jpg

 

[Svartbjørn]

Yes, I couldn't see the resemblance either!

It's actually a crop from an Alamy stock photo. You can even still see the watermarks on it.

View attachment 32999

http://c8.alamy.com/comp/C8EYKA/two...e-girls-one-blonde-one-brunette-uk-C8EYKA.jpg

Yanno, I DID see the A's but I dismissed them thinking it was something to fool facial recognition. Ive also never heard of Alamy or seen any of their work (that Im aware of) so might explain why it didnt trigger anything for me. Thanks for finding that guys.. howd you actually FIND the stock photo? The image searches I did here came up with nothing but Lotte Verbeek headshots, which is why I figured it was her. Ive not really seen her on TV or anything (dont watch much) so Im not all that familiar with her, which may also explain why I saw the resemblance but you guys didnt.

 

[Trailblazer]

howd you actually FIND the stock photo?

I used Yandex Image Search, which I find gives better results than Google or Tineye in many cases.

I just took a crop of the relevant area and used "search by image". It was the very first result in "Similar images":

 

[deirdre]

I used Yandex Image Search, which I find gives better results than Google or Tineye in many cases.

i forgot about Yandex.. good call. I just "save picture" from the twitter page and google only gave 2 hits. 1 trendolizer dot com for a comment she made about "gamer gate" and her twitter page.

but .. I don't know if it means anything, when I dragged the pic saved from twitter into google images the text they put in the search box was Finnish. (and "dirty" ie mild porn type stuff). Which I have no idea what that means.. if google did it or ... I don't know.

 

[marrowmonkey]

I used Yandex Image Search, which I find gives better results than Google or Tineye in many cases.

Thanks, very useful. After reading Svartbjørns post I also noticed the "a" watermark and even guessed it was an Alamy stock photo, but I couldn't find it on their site. No luck from Google either.

I did learn Google image search has an unofficial function that lets you search for matching faces though: just add &imgtype=face to the end of the search URL (that didn't help in this case either though).

 

[Svartbjørn]

I used Yandex Image Search, which I find gives better results than Google or Tineye in many cases.

I just took a crop of the relevant area and used "search by image". It was the very first result in "Similar images":

View attachment 33001

There ya go.. never heard of Yandex either.. Thanks Trailblazer.

 

[Agent K]

Does the fact that Yandex finds the profile pic suggest that the bot is NOT Russian, since a Russian troll would've known to check if Yandex finds the pic?

 

[deirdre]

Does the fact that Yandex finds the profile pic suggest that the bot is NOT Russian, since a Russian troll would've known to check if Yandex finds the pic?

No. There is still a copyright watermark across her face.

Anyone 'serious' would have spent 3 minutes and smudged the watermarks off. or used a pic with no watermarks.

 

[Agent K]

No. There is still a copyright watermark across her face.

Anyone 'serious' would have spent 3 minutes and smudged the watermarks off. or used a pic with no watermarks.

Do Russian trolls count as "serious"? If so, then all the more reason it's not a Russian troll bot.

 

[marrowmonkey]

No. There is still a copyright watermark across her face.

Anyone 'serious' would have spent 3 minutes and smudged the watermarks off. or used a pic with no watermarks.

Agreed, but on the other hand, if you are setting up thousands of bot accounts you might not have three minutes to spare. Might even be a fully automated process.

EDIT: But if they were 'serious' I suppose they wouldn't have used watermarked stock photos to begin with, at the very least purchased a bunch of royalty free pictures.

 

[Svartbjørn]

No. There is still a copyright watermark across her face.

Anyone 'serious' would have spent 3 minutes and smudged the watermarks off. or used a pic with no watermarks.

Do Russian trolls count as "serious"? If so, then all the more reason it's not a Russian troll bot.

Agreed, but on the other hand, if you are setting up thousands of bot accounts you might not have three minutes to spare. Might even be a fully automated process.

I think all three of you guys are saying the same thing.. That it IS most likely a bot (or a Russian Bot) of some kind -because- they didnt take the time to remove the water marks, or that its an automated process.

 

[Mick West]

That it IS most likely a bot

It's obviously a bot based on the repetitive content. The question is what is its intended purpose? It's a very simple bot, but someone has deliberately chosen the randomized content lists to make it seem (mostly) like a caricature of a snowflake liberal. Is it just some random experiment, of did something think it would have some effect?

 

[marrowmonkey]

It is really simplistic and something equivalent could probably be put together by most programmers over a weekend or so. If it is based on a generic bot program the owner might not even have to do any programming, only set up an account for it and provide some keywords. The fact it used stock photo image with watermarks indicates that whoever made it wasn't willing to put much effort or money into it either. I tried searching for some of the odd phrases it uses and did not find any other accounts that use the same keywords. So there does not appear to be multiple bots that use the same set of keywords.

It is only following 'marketing accounts'. That could actually be its real purpose. According to this NYT article some are willing to pay for fake twitter followers in order to make their account seem more popular:

Devumi sells Twitter followers and retweets to celebrities, businesses and anyone who wants to appear more popular or exert influence online. Drawing on an estimated stock of at least 3.5 million automated accounts, each sold many times over, the company has provided customers with more than 200 million Twitter followers, a New York Times investigation found.

https://www.nytimes.com/interactive/2018/01/27/technology/social-media-bots.html

It is probably obvious it is a bot to anyone who knows a bit of programming and is used to browse the world wide web. Is it obvious to everyone though? (I can not help to wonder if the goal could have been for people to notice that it is a bot and discuss it, just like we are doing now. Fake twitter follower seems more likely though).

 

[Trailblazer]

It is only following 'marketing accounts'. That could actually be its real purpose. According to this NYT article some are willing to pay for fake twitter followers in order to make their account seem more popular:

I'm not familiar with the alogrithms used to assess Twitter popularity, but do "active" followers (such as this bot, which tweets a lot) count for more than dormant followers? It could be as simple as that - create bots that would pass for human tweeters on the most cursory automated inspection.

 

[deirdre]

I'm not familiar with the alogrithms used to assess Twitter popularity, but do "active" followers (such as this bot, which tweets a lot) count for more than dormant followers? It could be as simple as that - create bots that would pass for human tweeters on the most cursory automated inspection.

I agree. although some of it's "word list" isn't typical. abortion, transgender etc are pretty trendy words but "chemtrails" and "alex jones" ? (and whoever that other guy is the bot keeps mentioning). gotta wonder where it got it's word list.

 

[marrowmonkey]

I agree. although some of it's "word list" isn't typical. abortion, transgender etc are pretty trendy words but "chemtrails" and "alex jones" ? (and whoever that other guy is the bot keeps mentioning). gotta wonder where it got it's word list.

Indeed, if you make a fake follower bot wouldn't you want it to avoid attention? This bot seems to do the opposite: it picks its topics from a list of politically contentious subjects.

Also, if a company was behind the bot and they had thousands, or even millions, of bot accounts I would expect they put more effort into the bot program. They could follow each other and retweet things occasionally for example. And they would probably reuse some of the same vocabulary since it would be too laborious to write unique dialogue for each of them.

This bot seems to be the only one of its type that I could find (my google-fu might be a bit rusty though), and It's the only one that use expressions like e.g. "and that's smashing" in that manner (that I could find).

Twitter might decide to change its policy on bots in the future. There are interesting, fun and maybe even useful bots; but it seems like the majority are taking up a lot of resources only to game the system in a detrimental way.

 

[deirdre]

It seems more like a kid making the account as a goof to me. It's too obviously a bot. I noticed the "a" s on her face even before I read any of the tweets.

It also could be one of those 'satire/troll/skeptic' type people who just like to make people look dumb by falling for obvious bunk.

could be anything.

 

[Hierophant]

I also think it's a hobby programmer playing around. The more "serious" bots/AIs appear to fulfill functions like enraging people, looking like they're winning a discussion regardless of content, making fringe groups look bigger, or discouraging people from bothering to answer to bunk by wearing them down in pointless discussions. This seems to do none of them. I guess you could just try to contact the owner of the twitter account by talking to the bot? Maybe the solution is as easy as asking just them directly what they're up to.

 

[Mick West]

I also think it's a hobby programmer playing around. The more "serious" bots/AIs appear to fulfill functions like enraging people, looking like they're winning a discussion regardless of content, making fringe groups look bigger, or discouraging people from bothering to answer to bunk by wearing them down in pointless discussions. This seems to do none of them. I guess you could just try to contact the owner of the twitter account by talking to the bot? Maybe the solution is as easy as asking just them directly what they're up to.

They have DMs disabled.

 

[Mick West]

I've been noticing a few more bots crop up, because they retweet things I've posted about chemtrails and other conspiracy theories like HAARP. Seem to be largely far right wing themed.

https://twitter.com/Am3rican_Only
https://twitter.com/RightsReBirth
https://twitter.com/NoMoreIllegals3
https://twitter.com/Make_US_Great76
https://twitter.com/SafeSchoolZone_
https://twitter.com/GunSafetyFacts

The last two are gun rights themed, and have retweeted stuff about the Vegas shooting. But the other are far more extreme. Like this:

 

[Ray Von Geezer]

This kind of reminds me of what happened with Tay - https://www.theverge.com/2016/3/24/11297050/tay-microsoft-chatbot-racist

Maybe AI is just predisposed to controversy and we're all doomed?

Ray Von

 

[Critical Thinker]

Perhaps related, from the NYTimes: Russian Trolls Used Vaccine Debate to Sow Discord, Study Finds

Excerpt

"You see this pattern," said David A. Broniatowski, a computer engineer at George Washington University and lead author of the study, which was published Thursday in the American Journal of Public Health. "On guns, or race, these accounts take opposite sides in lots of debates. They're about sowing discord."


With colleagues at the University of Maryland and Johns Hopkins University, Dr. Broniatowski looked at 899 vaccine-related tweets sent from mid-2014 to late 2017.

Some came from accounts known to send out spam or link to malware; more came from accounts that Congressional investigators and NBC News have identified as belonging to Russian trolls.


While the spammer and malware accounts mostly disseminated anti-vaccine messages, the Russia-linked ones played both sides.

Most of the anti-vaccine tweets repeated well-known but long-discredited rumors, such as those that vaccines cause autism or contain dangerous amounts of mercury. Others accused pharmaceutical companies of caring only about profits, not children.

Pro-vaccine tweets from the same accounts argued that vaccines saved lives. Some said they should be mandatory. Some were insulting, such as "You can't fix stupidity. Let them die from measles, and I'm for #vaccination."

But the Russians sometimes misread their audience, Dr. Broniatowski said, sending tweets that "didn't quite make sense, given the way Americans usually argue about vaccines."

Some, for example, suggested that God opposed vaccination. "I don't believe in #vaccines I believe in God's will," one read.

Additionally from CNN: Why Russian trolls stoked US vaccine debates

In the meantime, however, she said that the new study on Russian trolls meddling in US online vaccine debates is an example of how there has been growing distrust in science and public health initiatives, such as those underlying vaccinations.
"Both real people and trolls are capitalizing on that mistrust to push conspiracy theories out to vulnerable audiences," DiResta said.
"This isn't just happening on Twitter. This is happening on Facebook, and this is happening on YouTube, where searching for vaccine information on social media returns a majority of anti-vaccine propaganda," she said. "The social platforms have a responsibility to start investigating how this content is spreading and the impact these narratives are having on targeted audiences."

And from the WashingtonPost: Russian trolls and Twitter bots exploit vaccine controversy

These trolls may be flitting between divisive topics. Renee DiResta, who studies computational propaganda with the grass-roots collective Data for Democracy, said her own research suggests Russian trolls have fixated on the vaccine debate not to seed a public health crisis but to exploit its divisiveness.

Tweets she discovered from September 2017, for example, connect vaccine denial to racial anxieties: "Diseases Expert Calls for White Genocide Since Most Vaccine Deniers are White," several Russian trolls tweeted.

"It's opportunism — opportunistically amplifying controversial topics," DiResta said.

She pulled up one Russian bot, @WadeHarriot, that bounced around between various themes — anti-gay, pro-Ted Cruz, anti-Obama or pro-Trump until in early 2017 it began spouting anti-vaccination ideas. Only 38 of that user's 6,000 tweets were about vaccines. They began right around the time that the topic surfaced in the general news cycle because vaccine critic Robert Kennedy Jr. announced that President Trump asked him to head a new commission on vaccines. That commission never came into being.

 

[DavidB66]

View attachment 32926
The twitter account @jenn_m_tweets is a bot that randomly tweets every few minutes a simple opinion randomly constructed along the lines of:

"I think that (contentious subject) is (positive/negative) and (that's good/bad)

…..

What is the bot for? Is it some social experiment? Part of some nefarious Russian active measure to increase dissent? Just spam to make you click on associated links? (it never posts links). It follows mostly German Tweeters, which might be a clue. But I really can't figure out what it might be for.

The Times (UK) today (6 October 2018) has a front page story and an inside-page 'investigative report' with the headline 'Teenagers targeted by Russian trolls'. The front page article is here but may be paywalled for non-subscribers:
https://www.thetimes.co.uk/edition/news/uk-teenagers-targeted-by-russia-trolls-szzd5gd56

The gist of the report is that Russian state agencies are responsible for social media campaigns aimed at young people in the West (specifically in the UK) to sow fear and distrust. A wide range of issues are covered by the alleged Russian campaigns, some of them specifically of benefit to Russian interests (e.g. Syria, the Salisbury poisonings, and anti-fracking), but also some (e.g. vaccines, water fluoridation, and feminism) which seem aimed at stirring up dissent and anxiety in general. My eye was caught especially by the claim that 'Kremlin accounts promoted conspiracy theories as well. Some claimed that 'chem-trails' were sprayed into the atmosphere by aircraft to spread chemicals for sinister purposes. Others promoted the theory that Earth was flat'.

The source for all these claims is not always clear. Some of them are said to be based on media analysis commissioned by the Times from 'Renee DiResta, a specialist in disinformation who has advised the US Congress'. Other parts appear to be based on the newspaper's own research. I am bound to say that in such matters I always assume that the Times has at least consulted its contacts in UK intelligence services, if the stories are not actually planted by them. But perhaps this makes me a conspiracy theorist myself.

 

[scombrid]

I'm not laying all the conspiracy theory trolling and disinformation at the feet of the Russians.

Breitbart, Infowars, etc... have used conspiracy theories for fear mongering and a campaign to paint non-conservatives as some kind of global cabal in their anti-establishment populist movement. We've got homegrown disinformation bots and operatives on social media and such. NPR had an article about Kavanaugh this morning and a majority of the 1000s of comments were fake accounts. A notable was a facebook profile for Peter Griffin (family guy character) that was posting a response to almost every liberal comment in a thread that was 1000s of posts long.

 

[deirdre]

I'm not laying all the conspiracy theory trolling and disinformation at the feet of the Russians.

Breitbart, Infowars, etc...

because non-conservatives never troll or spread disinformation, they only participate in "satire".

The author of these tales is Christopher Blair. His website, TheLastLineOfDefense.org, is among the biggest sources of fake news online. Blair operates under a series of pen names — Ezekiel Wilekenmeyer, Christopher Lyman, Stryker, Busta Troll — but PolitiFact, which has been debunking his stories for months, pieced together his true identity.

In a series of email and Facebook chat interviews, Blair discussed the details of his brand of political parody.

Blair, who insisted that every word in our interviews was true, says he designed his parody site to fool conservatives with fake stories most people would find absurd.

https://www.politifact.com/punditfa...re-fooled-by-fake-news-this-man-probably-wro/

 

[scombrid]

because non-conservatives never troll or spread disinformation, they only participate in "satire".

I never said they didn't.

I just pointed out some very large organized efforts that are the modern equivalent of The Heartland Institute creating the Rush Limbaugh disinfo machine on AM radio in the 1980s and 1990s in the pre-internet days.

Like I wrote, Russians aren't the only ones creating the bots. They aren't even creating the content most of the time.

The author of these tales is Christopher Blair. His website, TheLastLineOfDefense.org, is among the biggest sources of fake news online. Blair operates under a series of pen names — Ezekiel Wilekenmeyer, Christopher Lyman, Stryker, Busta Troll — but PolitiFact, which has been debunking his stories for months, pieced together his true identity.

In a series of email and Facebook chat interviews, Blair discussed the details of his brand of political parody.

Blair, who insisted that every word in our interviews was true, says he designed his parody site to fool conservatives with fake stories most people would find absurd.

https://www.politifact.com/punditfa...re-fooled-by-fake-news-this-man-probably-wro/

That is certainly trolling but it is in a different family of activity than what Bannon and Alex Jones are doing along with Heritage and Heartland Institute.

Interesting to see who is sharing and spread his fake content.

 

[deirdre]

That is certainly trolling but it is in a different family of activity than what Bannon and Alex Jones are doing along with Heritage and Heartland Institute.

spreading fake content and disinformation is spreading fake content and disinformation. intentions are irrelevant.

But either way, this thread is about bots. and my point is we have no idea if these bots are Russian, conservative or liberals who think they are being funny.

 

[scombrid]

Yes this thread is about bots. Bots are re-broadcasting the fake content. Intentions of the content producers are relevant though. There is going to be a difference in persistence and sophistication when comparing a one-off troll and a disinformation campaign. It is true with the bots too and that is part of the reason for this thread. Knowing the motivation informs about who is doing it and how to deal with it.

 

[deirdre]

There is going to be a difference in persistence and sophistication when comparing a one-off troll and a disinformation campaign

she doesn't look very sophisticated to me.

 

[Jay Reynolds]

I happen to be acquainted with someone who studies bot networks as part of a .gov program and here is what they said:

This is obviously a bot yes. Here's a site that can produce a reliable 'Bot or Not' score for any twitter user. For this user it gave her a 2.1/5 which doesn't really qualify as a bot (4+ is likely a bot)
As humans we can see that it obviously is, but for software it's really difficult. This site looks at a users friends and followers and their tweets to determine if it is a bot.
On the site: enter a twitter handle, then check user, then followers. Quite a few of her followers are flagged as bots.
https://botometer.iuni.iu.edu/

To really dive in to this you need a network visualizer like Ora, Maltego, or Gephi.
This is Gephi: https://interworks.com/wp-content/uploads/2017/09/fb_network.jpg
With this Twitter user's Friends and Followers and their tweets, and maybe their Friends and Followers and tweets you can build a big network and easily see connections. Say for instance that a large majority or them make Alt-right posts, or they all suspiciously follow a couple Russian accounts. These sort of connections are hard to find but could build a more full picture of what is particular bot is up to.

 

[Sarah T.]

My experience with a Facebook "botnest"

I have had to learn to accept some questions will be unanswered. Mick starts this thread wanting to know "what is this bot doing?"

Sometimes you just can't know for sure. My guess, without digging into it- is that the bot is in a holding pattern- staying warm until it's pressed into service where it might be hindered by a gap in posting history. I think its activity is designed to walk the line between fooling bot detection and fooling a human who might have a chance encounter with it

JenniferBot's commenting reminded me of a cluster of bots I worked on with a lady who runs the FB page called "Unfakery"

Facebook livestream intro to the botnest:
https://www.facebook.com/unfakery/videos/173946579980773/

Followup on youtube Unfakery Video (first 7 minutes is on this topic)

She tends toward political subjects while I try to avoid them if at all possible- but sometimes we can combine our skills and this was really cool how it went down. Even though we dug up a ton of stuff- and even with the Craig Silverman's help (and he even had a direct line to FB) we never were able to trace this to its source.

She found a profile making strange comments. I assumed that if there was one there'd be many. I pulled up a ton more of these fake accounts tightly interconnected through friending, sharing each other's posts, sharing same links/ same format and commenting on each other's posts. I made a spreadsheet documenting all the nonsensical automated comments I could find. These comments were recycled by multiple profiles the comments had nothing to do with the post they responded to. As in the samples Mick gave- many comments had a "Mad Lib" automated structure- random preprogrammed filler words and phrases from list A and B to plugged in to make a variety of grammatical bot sentences. Not quite AI... a much "stupider" type of machine...just an autoposting program.

My guess: the comments had no purpose other than to amplify the posts of the friendBots- and to create a "chatter" of behavior between these profiles that mimicked the behavior of real human friends.
This faux social facade is like false fronts of a western movie town - looks authentic from one perspective- even more-so when projected on a screen, but falls apart if you bother to open even one door.

Unfakery hosts "live troll hunts" weekly on the FB page and those are also shared on the youtube channel (without the livestream commenting) They are not polished or scripted and are only directed by loose themes. This one, for the first nine minutes or so- shows a few of the remaining botnest profiles before they were finally taken down by Facebook. In the video notes there is a link to the spreadsheets we'd made as well as the Buzzfeed article that came from this (the article turned out to be more about Unfakery than the botnest- but very nice press for Felicia!)

 

[Mick West]

I've been noticing a few more bots crop up, because they retweet things I've posted about chemtrails and other conspiracy theories like HAARP. Seem to be largely far right wing themed.

https://twitter.com/Am3rican_Only
https://twitter.com/RightsReBirth
https://twitter.com/NoMoreIllegals3
https://twitter.com/Make_US_Great76
https://twitter.com/SafeSchoolZone_
https://twitter.com/GunSafetyFacts

The above six have all been suspended. I wonder if they eventually were "activated" after building up their Twitter cred and then were suspended.

I revisited this because I saw a few more over the last couple of days, all retweeting Metabunk stuff:

https://twitter.com/WillsDoctorate
https://twitter.com/UKCTJanedoe
https://twitter.com/CanadaAyy
https://twitter.com/safetysaraa
https://twitter.com/PreventGunz - Pushing illegal guns is one of this generations most pressing matters! Newark, DE Joined Mar 2019
https://twitter.com/HowToStpDem - We have to stop the march of the NRA for the sake of our communities! Delaware, USA Joined Mar 2019
https://twitter.com/S0ccerDayz
https://twitter.com/NoGunsNoMo
https://twitter.com/superbowldanger - Super Bowl - the best day of the year. Oakland, CA Joined Jun 2019
https://twitter.com/Thesp0rtszguys - Thesp0rtszguys works to bring the love of sports to the internet. Promoting the GooDOle bar with patrons and the like! Santa Monica, CA Joined Mar 2019


Following a similar pattern to before. The accounts are either framed as politically anti-gun, or are generic single interest, like sports or cars, or they claim to be interested in understanding some aspect of a politically divisive issue, like mass shootings, or terrorism.

They have appropiate header and background images, the tag lines often have a non-English feel to them:

What I'm wondering is what Twitter is doing here? It seems very obvious that these are bots, perhaps with some human intervention in the setup. Is Twitter simply incapable of detecting them? Or are they flagged and then blocked if they step over the line. If the latter, then why are they being kept around?