6 ms·

I have a few questions that may have some reasonable answers, but would hopefully spur some thoughtful discussion. 1) What is a "troll farm" and what is the cr

I have a few questions that may have some reasonable answers, but would hopefully spur some thoughtful discussion. 1) What is a "troll farm" and what is the criteria used for a distinction between a troll farm and merely some points of view one might disagree with? Maybe this is mentioned somewhere, but I don't see it clearly defined in the link. 2) What is the proof that these are indeed foreigners? It would be trivial for a social media company to label domestic users they find distasteful as foreigners to more easily silence certain points of view, especially given the media's unwillingness/inability to understand nuanced technology issues. 3) Heck, where is the proof that these pages are even moderately operated by actual people and are not mostly bot-generated content? 4) We live in a global, connected world that embraces diverse points of view, right? Are eastern Europeans or others, many of whom still embrace a traditional Christian heritage, expected to not post about their point of view because somebody in San Francisco claims its a troll? 5) How can a data scientist make a definitive claim that the people running these sites have "never been to church"? Maybe these social media firms are making big assumptions about people and are entirely wrong. Maybe they have entirely far too much information about people.

mistermann 5y ago
My question is whether the location could be spoofed using a VPN or something like this...could it not?
- marcus_holmes 5y ago
  Pretty sure FB doesn't rely on just IP address to locate its users.
  - mistermann 5y ago
    I am interested in the means by which they DO identify user location, and whether the ones they use can be spoofed. And for those who cannot help but ask "But why would....", while that is an interesting and valid question, it is orthogonal to this specific question.
    - marcus_holmes 5y ago
      Well, if the app is installed, then there's a gajillion methods (GPS, Wifi, etc) For conventional browsers in a desktop environment I'd suspect all the normal methods of fingerprinting browsers would reveal a lot. And Netflix et al can detect VPNs from a browser, so clearly that's not impossible. I notice that Google Maps on my browser knows where I am despite me using a VPN, but then I'm also logged in to Google on my phone, so maybe it's just tracking my phone location rather than my browser location. Interesting question - anyone know anything definite?
- p49k 5y ago
  Why would 19 of the top 20 Christian pages be run by Americans who specifically only connect to VPNs in Kosovo and Macedonia when they manage their pages? I think at some point common sense comes into play here.
weswpg 5y ago
Content that is inauthentic and/or misrepresenting the origin / source is definitely troll material
mcphage 5y ago
> 1) What is a "troll farm" and what is the criteria used for a distinction between a troll farm and merely some points of view one might disagree with? Maybe this is mentioned somewhere, but I don't see it clearly defined in the link. > 3) Heck, where is the proof that these pages are even moderately operated by actual people and are not mostly bot-generated content? Bot generated content would definitely qualify as a troll farm.
p49k 5y ago
From the study: 1) Troll farm = Professionalized groups that work in a coordinated fashion to post provocative content 2) Through internal Facebook research in which documentation was leaked to journalists. 3) That's irrelevant to the story. The point is that the claimed motives behind the operators of the page are inauthentic and dishonest, and those claims could apply to either/both human and bot-generated content. 4) This question just makes your entire post sound like disingenuous trolling, but people of course expect honesty, and a page called "Kosovo engages with American Christians for profit" would not be criticized in the same way. 5) That's an obvious strawman - the people running the pages have shown to be dishonest in every way that their honesty can reasonably be measured. https://s3.documentcloud.org/documents/21063547/oct-2019-facebook-troll-farms-report.pdf https://s3.documentcloud.org/documents/21063547/oct-2019-fac...
- logicalmonster 5y ago
  > 1) provocative content You know, defining "provocative content" is a subjective endeavor. What is viewed as normal conversation in the real world by most people might be viewed as extremely problematic, unmitigated, obvious hate-speech by some in a Silicon Valley thought-bubble. How do we really clarify this? > 2) internal Facebook research in which documentation was leaked to journalists. a) How do we know that some Facebook employees are not using their own biases to give incorrect subjective labels to some content based on their own activist viewpoints? Maybe, maybe not, but I would stress that we're relying on subjective interpretations here and just have to trust them. b) How do we know that the journalists who are interpreting this data and reporting on it are both willing and able to really do anything other than take the face-value interpretation of the data they were handed? I can count the number of journalists I've seen in my life who tried to dig into the nuanced details behind stories on one hand. > 3) the claimed motives behind the operators of the page are inauthentic and dishonest This is a hypothesis which may be (or likely is) reasonably supportable, but I would clarify that mind-reading has never been a thing. > 4) a page called "Kosovo engages with American Christians for profit" would not be criticized in the same way. Maybe. My goal isn't to necessarily defend these website operators. They might be, or even most likely are, creating random content to achieve fame and/or profit through social media. Ok, does that separate them from 99.9% of people/businesses/organizations on social media? > 5) the people running the pages have shown to be dishonest in every way that their honesty can reasonably be measured. My main goal in this set of comments isn't to necessarily defend the honor of these website operators. Maybe they're pure, manipulative scum? My main goal is to point out that Facebook (and the media) has shown to be dishonest in every way that their honesty can reasonably be measured. I do not understand how anybody can trust this interpretation of the facts at face value for many, many reasons here.
marcus_holmes 5y ago
I'm no expert, but I'll take a shot at some answers... > 1) What is a "troll farm" and what is the criteria used for a distinction between a troll farm and merely some points of view one might disagree with? Maybe this is mentioned somewhere, but I don't see it clearly defined in the link. It's down to intent. I'd define it as: "A troll farm is intended to manipulate opinion in some manner for political goals by (mis)representing itself as a genuine member of the community it intends to manipulate." > 2) What is the proof that these are indeed foreigners? It would be trivial for a social media company to label domestic users they find distasteful as foreigners to more easily silence certain points of view, especially given the media's unwillingness/inability to understand nuanced technology issues. I don't think TFA provides any proof of its claims, but points to a MIT Technology Review article [0] that claims a report "obtained by MIT Technology Review from a former Facebook employee not involved in researching it" as its source. It would be entirely reasonable to dismiss the entire article as "fake news" based on this attribution. > 3) Heck, where is the proof that these pages are even moderately operated by actual people and are not mostly bot-generated content? Again, no proof provided. Does that make a difference to the intent of the page, though? > 4) We live in a global, connected world that embraces diverse points of view, right? Are eastern Europeans or others, many of whom still embrace a traditional Christian heritage, expected to not post about their point of view because somebody in San Francisco claims its a troll? That's not what these pages do (post content in Eastern European languages intended for Eastern European audiences addressing issues faced by Eastern European christians). They're posting in English about things that are contentious for American readers. > 5) How can a data scientist make a definitive claim that the people running these sites have "never been to church"? Maybe these social media firms are making big assumptions about people and are entirely wrong. Maybe they have entirely far too much information about people. I didn't read that as a definitive claim, but more as an opinion expressed by one person. [0] https://web.archive.org/web/20210917012652/https://www.technologyreview.com/2021/09/16/1035851/facebook-troll-farms-report-us-2020-election/ https://web.archive.org/web/20210917012652/https://www.techn...
photochemsyn 5y ago
I would just note that an 'advertising agency' is rather indistinguishable from a 'troll farm' - using provocative content to get eyeballs on their product is basically the whole game.

mistermann 5y ago

My question is whether the location could be spoofed using a VPN or something like this...could it not?

marcus_holmes 5y ago
Pretty sure FB doesn't rely on just IP address to locate its users.
- mistermann 5y ago
  I am interested in the means by which they DO identify user location, and whether the ones they use can be spoofed. And for those who cannot help but ask "But why would....", while that is an interesting and valid question, it is orthogonal to this specific question.
  - marcus_holmes 5y ago
    Well, if the app is installed, then there's a gajillion methods (GPS, Wifi, etc) For conventional browsers in a desktop environment I'd suspect all the normal methods of fingerprinting browsers would reveal a lot. And Netflix et al can detect VPNs from a browser, so clearly that's not impossible. I notice that Google Maps on my browser knows where I am despite me using a VPN, but then I'm also logged in to Google on my phone, so maybe it's just tracking my phone location rather than my browser location. Interesting question - anyone know anything definite?
p49k 5y ago
Why would 19 of the top 20 Christian pages be run by Americans who specifically only connect to VPNs in Kosovo and Macedonia when they manage their pages? I think at some point common sense comes into play here.

weswpg 5y ago

Content that is inauthentic and/or misrepresenting the origin / source is definitely troll material

mcphage 5y ago

> 1) What is a "troll farm" and what is the criteria used for a distinction between a troll farm and merely some points of view one might disagree with? Maybe this is mentioned somewhere, but I don't see it clearly defined in the link. > 3) Heck, where is the proof that these pages are even moderately operated by actual people and are not mostly bot-generated content? Bot generated content would definitely qualify as a troll farm.

p49k 5y ago

From the study: 1) Troll farm = Professionalized groups that work in a coordinated fashion to post provocative content 2) Through internal Facebook research in which documentation was leaked to journalists. 3) That's irrelevant to the story. The point is that the claimed motives behind the operators of the page are inauthentic and dishonest, and those claims could apply to either/both human and bot-generated content. 4) This question just makes your entire post sound like disingenuous trolling, but people of course expect honesty, and a page called "Kosovo engages with American Christians for profit" would not be criticized in the same way. 5) That's an obvious strawman - the people running the pages have shown to be dishonest in every way that their honesty can reasonably be measured. https://s3.documentcloud.org/documents/21063547/oct-2019-facebook-troll-farms-report.pdf https://s3.documentcloud.org/documents/21063547/oct-2019-fac...

logicalmonster 5y ago
> 1) provocative content You know, defining "provocative content" is a subjective endeavor. What is viewed as normal conversation in the real world by most people might be viewed as extremely problematic, unmitigated, obvious hate-speech by some in a Silicon Valley thought-bubble. How do we really clarify this? > 2) internal Facebook research in which documentation was leaked to journalists. a) How do we know that some Facebook employees are not using their own biases to give incorrect subjective labels to some content based on their own activist viewpoints? Maybe, maybe not, but I would stress that we're relying on subjective interpretations here and just have to trust them. b) How do we know that the journalists who are interpreting this data and reporting on it are both willing and able to really do anything other than take the face-value interpretation of the data they were handed? I can count the number of journalists I've seen in my life who tried to dig into the nuanced details behind stories on one hand. > 3) the claimed motives behind the operators of the page are inauthentic and dishonest This is a hypothesis which may be (or likely is) reasonably supportable, but I would clarify that mind-reading has never been a thing. > 4) a page called "Kosovo engages with American Christians for profit" would not be criticized in the same way. Maybe. My goal isn't to necessarily defend these website operators. They might be, or even most likely are, creating random content to achieve fame and/or profit through social media. Ok, does that separate them from 99.9% of people/businesses/organizations on social media? > 5) the people running the pages have shown to be dishonest in every way that their honesty can reasonably be measured. My main goal in this set of comments isn't to necessarily defend the honor of these website operators. Maybe they're pure, manipulative scum? My main goal is to point out that Facebook (and the media) has shown to be dishonest in every way that their honesty can reasonably be measured. I do not understand how anybody can trust this interpretation of the facts at face value for many, many reasons here.

marcus_holmes 5y ago

I'm no expert, but I'll take a shot at some answers... > 1) What is a "troll farm" and what is the criteria used for a distinction between a troll farm and merely some points of view one might disagree with? Maybe this is mentioned somewhere, but I don't see it clearly defined in the link. It's down to intent. I'd define it as: "A troll farm is intended to manipulate opinion in some manner for political goals by (mis)representing itself as a genuine member of the community it intends to manipulate." > 2) What is the proof that these are indeed foreigners? It would be trivial for a social media company to label domestic users they find distasteful as foreigners to more easily silence certain points of view, especially given the media's unwillingness/inability to understand nuanced technology issues. I don't think TFA provides any proof of its claims, but points to a MIT Technology Review article [0] that claims a report "obtained by MIT Technology Review from a former Facebook employee not involved in researching it" as its source. It would be entirely reasonable to dismiss the entire article as "fake news" based on this attribution. > 3) Heck, where is the proof that these pages are even moderately operated by actual people and are not mostly bot-generated content? Again, no proof provided. Does that make a difference to the intent of the page, though? > 4) We live in a global, connected world that embraces diverse points of view, right? Are eastern Europeans or others, many of whom still embrace a traditional Christian heritage, expected to not post about their point of view because somebody in San Francisco claims its a troll? That's not what these pages do (post content in Eastern European languages intended for Eastern European audiences addressing issues faced by Eastern European christians). They're posting in English about things that are contentious for American readers. > 5) How can a data scientist make a definitive claim that the people running these sites have "never been to church"? Maybe these social media firms are making big assumptions about people and are entirely wrong. Maybe they have entirely far too much information about people. I didn't read that as a definitive claim, but more as an opinion expressed by one person. [0] https://web.archive.org/web/20210917012652/https://www.technologyreview.com/2021/09/16/1035851/facebook-troll-farms-report-us-2020-election/ https://web.archive.org/web/20210917012652/https://www.techn...

photochemsyn 5y ago

I would just note that an 'advertising agency' is rather indistinguishable from a 'troll farm' - using provocative content to get eyeballs on their product is basically the whole game.