Hard News by Russell Brown

One other thing that could be relevant: Lucy Lawless retweeted my unkind tweet about Ben Carson being stupid this week. My tweets usually get 1000-5000 impressions. That one is closing in on 40,000.

As of last week one can block egg's, so their replies don't show up in your mentions/timeline - it seems this feature was long requested, so hopefully its of some use managing what looks to be quite a nuisance:

https://thenextweb.com/twitter/2017/03/01/twitter-now-lets-auto-mute-eggs-sketchy-accounts/#.tnw_4Xh77mxy

I gather there are also ways to remove them en-masse ,,,

In fairness, though, it was a great jab and deserved the coverage it got.

Perhaps as bots get more prevalent and realistic, they'll undermine the (flaky) economic basis of social media - will companies pay to advertise to an audience of non-consuming computer programs?

Perhaps as bots get more prevalent and realistic, they’ll undermine the (flaky) economic basis of social media – will companies pay to advertise to an audience of non-consuming computer programs?

The rampant fakery of the internet advertising market hasn't killed off "programmatic" yet – agencies understandably don't want to tell clients that most of it's bullshit.

But it may only be a matter of time.

In fairness, though, it was a great jab and deserved the coverage it got.

Aw, thanks!

So far, there’s been a cluster of features that together have formed a reliable diagnostic for the recent flurry of bot-posts here:
(i) new username;
(ii) random selection of thread (often long-dormant, because most threads are);
(iii) lack of new relevant content (though some superficial relevance may be achieved through copying key words and phrases from the thread, sometimes with some additional generic response, e.g. that the topic “is interesting”);
(iv) nonnative grammar or vocabulary selection;
(v) inclusion of a link to a website offering some service (either as in-text link, or as an icon in the message header).

But none of these features by itself would be a sufficient diagnostic; and one may expect some improvements in (ii), (iii), and (iv), possibly to such an extent that it would be hard to tell a bot from a genuine commenter.

But whatever it is, it feels like something is going on. And the fact that that something includes a ramping-up of anti-Trump provocations makes it even more intriguing. Is this a new entrant to the mass-manipulation game? Or is the same old network of automated emotional dividers recalibrating for a new zeitgeist?

With the emergence of Robert Mercer and Cambridge Analytica, Rupert Murdoch and Crosby/Textor already look like relics of the last decade.

On Twitter, Sleeping Giants and Shannon Coulter have been doing some sterling work in defunding and boycotting sites like Breitbart, one tweet at a time. But boycotts can only go so far - if we really want to take down Mercer and his ilk, maybe Al Capone's one fatal weakness comes to mid.

Part of me is: that's a bit strange.
But most of me is: is this some kind of "twitter circle-jerk" meaning nothing at all?
(not meaning to be rude)

That said, seems like there's little point in arguing on twitter.
Nobody goes there to get their mind changed ...

Out damned bot!
How long does a bot 'live'?
Once released can they be controlled/instructed (or taught or misled)?
...or are they autonomous thereon in?
Can they 'reproduce'?

I've noticed (and reported) the waves of dubious looking posts that linger refers to - they are getting better at appearing 'real' ...

they are getting better at appearing 'real' ...

Jürgen Tomicek.

they are getting better at appearing ‘real’

They sure are. Including an apparently positive if bland reaction is scarily effective social engineering -- you want to give those statements of approval the benefit of any doubt, even when you know you're being manipulated. It's the selection of dormant threads that is the best indicator that an automated process, rather than a live human, is involved. If there hadn't also been such clear examples resuscitating years-old threads, I would have thought twice about reporting the ones attached to merely month-old threads.

Dark data...
With the bot posts here, what is the aim?
Am I supposed to click on the link or website, will that then give 'them' a chance to 'infect' my computer via cookies, or somesuch, or does it show them a live address to target - like email spam?

What does this emerging life form want?
Can we co-exist and trade with them?

With the bot posts here, what is the aim?

It's simple. They're forum spammers aiming to place links on any site which rates well with Google.

I help to run a specialised forum which is only open to NZ and AU members. While we make this clear on the front page, we still get new signups most days from dodgy countries. Sometimes it's bots, other times it's actual people -- often from India or Russia. "Buy 1,000 Google 3+ links for $100" is their motivation. And as Linger observed above, reviving dormant threads is a useful method for these people as they have zero interest in intereacting with real people.

Some of the bots are easy to spot. We've set up our forums to ask for the user's "first name" during the signup process. Bots aren't that smart and almost always use their (randomly generated) username for this field. "Hi, my name is uoghqwfhjiP6$7 and I want to join your group." Ah huh.

Once they have a login, they can change their online signature (or their website link on PAS) at will. This allows the operators to change their links to suit whatever campaign they're being paid for. So suddenly you can have 1,000+ links from credible sites such as this, pointing at a new site which is selling "men's health" products or whatever. Those links boost the site's profile and increase its ranking in Google searches.

Some of the bots which have appeared on PAS recently have attempted to sign up on our forums at around the same time using the same website links. We blacklist vast IP ranges, check every new member and disable bots straight away. But still they keep coming. It's a game of whackamole.

However, that's forum spammers. Twitter bots are different.

For the Twitterbots, I would hypothesise that if your primary intent is to cause chaos in the U.S. then you have no particular loyalty to any faction in the US and will boost or attack on the basis of what seems to cause the most friction.

I haven't noticed any twitterbots yet. But I only follow and am followed by less than 400 people. My concept is to keep it sort of village sized. But my purpose is not to speak directly to vast numbers in the first place. I'd rather whatever goodness there is in anything I tweet be mediated by a small number and thus only ever go viral if it's really interesting/clever etc.

I presume once you start hitting city-state sized numbers of followers, you're really not using Twitter in a way that resembles how the majority use it.

The response to your St Petersberg jibe may not prove the bot is a real person. AI is getting better and even capable of appearing to pass the Turing Test.

Some of them are definitely human: one responded with telling (and amusing) indignation to my customary greeting, “How’s the weather in St Petersburg?”, demanding to know whether I was racist against Russians. You blew your cover there, dude.

Aw, damn… the only time I regret dropping the “report and block” hammer and missing that. Even then, they’re rather easy to detect – I wonder if they understand (or much care) it only takes a couple of clicks to reveal a string of identically worded replies obviously triggered by a keyword.

and even capable of appearing to pass the Turing Test.

" ...these aren't the bots you're looking for."

Re spying, technology and Trump links. This stuff that Matt Nippert is uncovering about Peter Thiel and his influence in NZ is very worrying.

Speculatively, if there was contact between the Trump campaign and people they shouldn't be talking to, and the people they shouldn't have been talking to see there being more value in creating chaos for the administration than supporting it, then they can leak contact information to the degree that suits them.

I am particularly musing about this in relation to today's Roger Stone says he was in contact with the DNC hackers but it was innocuous.

I've noticed this cluster of opinions:

* It's not that I support Trump but Hillary Clinton is a warmonger.

* It's not that I support Assad but there's no moderate opposition.

* It's not that I support Putin but he shouldn't be provoked.

Not sure what motivates the pretense. In the US some of that is you can earn money in the pundit world - being an arsehol gets attention. But there's a growing authoritarian voice out there on the internet that sees advantage in hiding behind the It's Not That... facade. And certain pundits are feeding that for their own reasons.

I've noticed this cluster of opinions:

* It's not that I support Trump but Hillary Clinton is a warmonger.

* It's not that I support Assad but there's no moderate opposition.

* It's not that I support Putin but he shouldn't be provoked.

Not sure what motivates the pretense. In the US some of that is you can earn money in the pundit world - being an arsehol gets attention. But there's a growing authoritarian voice out there on the internet that sees advantage in hiding behind the It's Not That... facade. And certain pundits are feeding that for their own reasons.

It's the same kind of fallacy that's closely related to saying, "some of my best friends are Jewish, but..." or saying, "I'm not anti-gay, but I don't approve of homosexual acts."

Or indeed the old standard
I'm Not Racist Butt
(isn't it dark up there)

Important article about the sinister Robert Mercer - the money behind Donald Trump and Cambridge Analytica who specialise in “election management strategies” and “messaging and information operations”.
https://www.theguardian.com/politics/2017/feb/26/robert-mercer-breitbart-war-on-media-steve-bannon-donald-trump-nigel-farage