> Even on HN the past few years the number of propaganda posts and comments has gone up and up and up. Continues to increase right now.
How do you know this? What do you use to decide? As far as I can tell, the way most people decide this is: if I don't like it, it's propaganda. Nobody thinks that consciously, of course, but it matches what people say surprisingly closely. That has no evidentiary value, except as an indication of what someone dislikes. You learn this very quickly if you're responsible for the community as a whole, because people have such different likes and dislikes that basically everything contentious gets dismissed as 'propaganda' or some such abuse.
Two things seem clear: (1) propaganda exists; (2) "I don't like it" is no way to identify it. There needs to be something more objective. When people are worried about abuse on HN and we look into it, our bar is that we have to see at least some evidence, somewhere, of something amiss. That's as low as the bar can get, yet even that is enough to reveal that way over 90%, probably over 99%, of perceptions of propaganda have no basis beyond the feelings of the perceiver.
Throwing some ideas here since it's a bit of a meta-discussion but on topic.
I think other social media sites have realized that there's more than "legitimate users" as well. Identifying propaganda is tricky and it certainly can't be just "because I don't like it" but there are some factors that can be checked:
- Accounts focused on one or few (thorny) themes (they might talk about other things for some "reputation laundering" but they're mostly focused on some specific subjects)
- Activity times not corresponding to the purported user location (and centered around "working hours").
- Linguistic patterns also not corresponding to the purported user location (or suggesting usage of automated translation).
I can't speak for the person you are replying to, but I already decided, so I don't need to show evidence for anything. I feel so much less guilty about all the time I waste here having total confidence that tier 1 state actors are in fact using their very best operatives and technology (not the low end crap they use to fool the rubes on Facebook, Youtube, and editors of newspapers of record) to wage a highly sophisticated influence operation against us. Surely if they could only succeed in brainwashing us, the users of HN, to go along with their dastardly plans the entire free world will collapse and world domination will be theirs.
So yeah whatever, say it's not happening, based on all the evidence that having complete access to the site gives you. I don't believe you. I just know, deep down, based on my limited unprivileged access and my casual analysis of a small part of the content, that anyone who disagrees with my fundamental beliefs on here is a total shill.
I'm reading that as satire (edit: though your last line was so uncannily realistic I really had to triple-check). To reply seriously to one bit: I'm not saying it isn't happening. I'm just asking the people who are sure it's happening, how they know. Either they have a way of deciding this that actually works, in which case we need to know what it is, so we can use it—or they don't, in which case the answers are interesting for other reasons.
I don't believe it's happening here unless we want to count the Rust Evangelism Strike Force. If it were then you would have noticed or people doing analysis on the public data would have posted reports of accounts involved and how their activity is coordinated. Look at all the posts over the years breaking down what gets to the front page and who posts it. It's not as if there aren't users who are actually paying attention. Actual campaigns have been active on Facebook, Twitter and Reddit, and indeed there are reports making specific and detailed accusations beyond "I don't like your comment history".
Joking aside, I believe that the minimum standard for a "influence operation" is where "operators" use a greater number of sock puppet accounts in a centrally coordinated manner to dominate the discussion forum. There has to be deception: pretending to be multiple people.
If I ever see activity that appears to meet that standard I'll report my evidence of it to the admins, trust they will take appropriate action if necessary, and continue to assume everyone who doesn't agree with me is a reasonable person acting in good faith.
Could you quantify some of your assertions, so I can better understand your perspective? I'm not looking for sources and serious analysis, but just looking for more precise numbers. What % of posts do you think are manipulated? How much worse has that number gotten over time?
I think state actors would be best served by making bots to upvote comments that serve their needs rather than directly commenting their position.
Ballpark, I'd estimate 1% of comments on sensitive topics, but 20% of votes on these topics are done with purpose of intentionally misleading the public.
It was satire, I was only kidding. As of this time I believe about 0% of posts here are manipulated. It's not to say we're not all exposed to propaganda and spin all the time, just that HN is not a forum that makes sense to target - it's too small.
Don't you know? Everyone on Hacker News is a bot. Even me. Even you.
\end{jest}
More seriously...
The thing about propaganda is that once someone starts believing it, that someone can potentially (and often does) spread it organically. Hacker News can inadvertently be a hotbed for propaganda spreading of all sorts without even a single bot or paid poster/commenter. I know I've fallen victim to that before, parroting something that ended up being misleading at best because I had been convinced through disinformation that it was the truth and therefore needed spread. I therefore tend to be cognizant of when others appear to be doing the same, seeing in them the same patterns I expressed.
In other words: just because it ain't violating the HN community guidelines doesn't mean it ain't propaganda.
Good question. I don't know if there are any objective criteria, given that the difference between a propaganda campaign and a more benign information campaign is often pretty blurry and subjective.
I think the best bet is to call out (or at least note) the common patterns as they pop up. For example, I know you happen to hold a dim view of accusing others of "whataboutism", but that does happen to be a very common form of modern-day organically-spreading propaganda ("Who are you to say my country does evil things when your country does evil things, too?", says the accidental propagandist, forgetting that two wrongs don't make a right and that one country doing evil things does not excuse another country from doing evil things), and it's important to be able to identify that for what it is. Granted, sometimes that line of thinking is often valuable in the reverse (i.e. "I don't think it's right that your country is doing this evil thing, so I shouldn't be okay with my country doing a similar evil thing."), but it doesn't seem like those sorts of comparisons are typically done with that sort of self-awareness or good-faith desire to understand the other side of the debate.
"Accidental propaganda" is an oxymoron. What's the line between that and just being wrong or disagreeing? It isn't bad arguments; people are even more likely to call 'propaganda' (or astroturfing, shilling, etc.) against good arguments, because good arguments one dislikes are even more activating than bad arguments one dislikes.
The problem is that there are two different phenomena surrounding those words. The first is real abuse; the second is people labeling comments they dislike as abuse because they can't imagine anyone could hold them in good faith. As far as we can tell, the second problem is much more widespread than the first.
As far as I can tell, the way most people decide this is: if I don't like it, it's propaganda
I don't think that "most" is accurate.
It is not hard to recognize propaganda, even when it coincides with one's established views.
In elementary school we had many quizzes and drills on this sort of stuff as part of our civics courses. The problem is, I doubt schools teach this stuff anymore, even though this kind of very basic critical thinking is very important for a functioning society.
I'm talking about the things people say on internet forums like HN about propaganda and related terms (astroturfing, shilling, brigading, foreign agents). At least on HN I can tell you that "most" is an understatement. That's plain from the comments and even more plain in the voting data. (I should make that more precise, though. Obviously I'm not reading people's minds to know why they say 'propaganda', etc.; it's rather that what people say about this is overwhelmingly correlated with whatever their view on the underlying issue happens to be.)
Elementary school is a pretty different context. Schoolchildren may be more reliable on this, actually. Although they tend to imbibe the views of their parents, they haven't yet had time to build up the same fixed emotional habits around them.
Those civic classes only prepared people for the 20th century world of propaganda comprised of overly-enthusiaatic pamphlets and loudspeakers, and to their credit did a good job of it. The problem is that those people are still unequipped to decide whether WeebFoxMaster13 is a real person or some marketer being paid $10 an hour in order to sell them on the idea of political activism
I don't think so. Propaganda works by convincing you without making you suspicious. If you can recognize it, it has already failed. Of course, not all propaganda is actually concerned with convincing you. Most of it is probably only concerned with convincing a majority, your minority opinion might not be worth the effort.
I think this hinges on one's definition of propaganda. I believe most propaganda starts as being spread by state actors - but the majority of the spread is probably done by normal citizens.
If an intelligence agency was directly posting false or misleading information on HN for the purposes of influencing readers, presumably if HN mods found evidence of this they would try to restrict it to the extent possible.
The much more challenging issue is that once a propaganda meme has taken hold in a normal citizen, they will also spread these ideas. This creates a nearly-impossible to moderate situation where a percentage of the population sincerely holds a belief and chooses to spread it.
For example: in recent threads about the release of Snowdens book, there were some comments that presented ideas as fact which are contradicted by publicly available evidence. I think it is fair to call this 'propaganda', in the sense that many of these ideas were likely deliberately crafted by intelligence orgs at inception, but these is no actionable moderation that can occur to prevent it. Impossible to tell if someone spreading an incorrect fact is doing so out of sincere belief or ulterior motive.
I agree that it seems like more propaganda has been appearing on HN in the past couple years, but sadly I think it is impossible to do anything to prevent it.
Any message board participant with even the most casual desire to score points and dunk on people could connect virtually any argument to "state-generated propaganda" using this logic.
How do you know this? What do you use to decide? As far as I can tell, the way most people decide this is: if I don't like it, it's propaganda. Nobody thinks that consciously, of course, but it matches what people say surprisingly closely. That has no evidentiary value, except as an indication of what someone dislikes. You learn this very quickly if you're responsible for the community as a whole, because people have such different likes and dislikes that basically everything contentious gets dismissed as 'propaganda' or some such abuse.
Two things seem clear: (1) propaganda exists; (2) "I don't like it" is no way to identify it. There needs to be something more objective. When people are worried about abuse on HN and we look into it, our bar is that we have to see at least some evidence, somewhere, of something amiss. That's as low as the bar can get, yet even that is enough to reveal that way over 90%, probably over 99%, of perceptions of propaganda have no basis beyond the feelings of the perceiver.