5 ms·

Thanks for the response. That first link is a Twitter bot account. There’s a bunch of them under different names. You’ll notice the #HackerNews hashtag, and al

Thanks for the response. That first link is a Twitter bot account. There’s a bunch of them under different names. You’ll notice the #HackerNews hashtag, and all the links are HN posted articles. Again, the same bots also seem to double duty product traffic spam. The third link’s relevance is that there is a very large body of evidence from respected institutions that social media is targeted by influence campaigns. I don’t know why anyone would think that the campaigns are targeting Facebook, Twitter, Reddit, but not HN. From that article “The platforms removed more than 317,000 accounts and pages from ‘cyber troops’ actors between January 2019 and November 2020.” Perhaps you guys don’t believe these studies? They aren’t always obvious things, that’s the whole point. For instance with Facebook foreign governments were pushing not just obvious political misinformation but trying to marginalize groups by making them seem less rational in online topics. Perhaps you might want to survey the HN community. I honestly believe that it’s become more toxic and divided, and it’s becoming clear to anyone paying attention that social media is being manipulated by large well funded groups. Again, not sure if your in a position to do anything about it. But maybe do some hard research in the area. It’s not “fanciful imagination” when there are studies from respected outlets. Here’s a starting point I’ve used for my own research: https://scholar.google.com/scholar?hl=en&as_sdt=0%2C44&q=social+media+manipulation&oq=social+media+man https://scholar.google.com/scholar?hl=en&as_sdt=0%2C44&q=soc... From one of these studies.. “Specifically, this study examines a collection of tweets relating to a much-publicized fan dispute over the Star Wars franchise film Episode VII: The Last Jedi. This study finds evidence of deliberate, organized political influence measures disguised as fan arguments. The likely objective of these measures is increasing media coverage of the fandom conflict, thereby adding to and further propagating a narrative of widespread discord and dysfunction in American society” And from another. “Despite being outnumbered by several orders of magnitude, just a few thousands of bots generated spikes of conversations around real-world political events in all comparable with the volume of activity of humans. We discover that bots also exacerbate the consumption of content produced by users with their same political views, worsening the issue of political echo chambers. ”

dang 5y ago
I have a bunch more to say about this but unfortunately I have to run out for a training. A couple super quick points though: > Facebook, Twitter, Reddit, but not HN HN is orders of magnitude smaller. Sites face qualitatively different problems at each order of magnitude. I'm not saying that proves anything, but it's a sensible reason why we don't see the data they do. > just a few thousands of bots generated spikes of conversations I can tell you from deep familiarity with HN's data that it isn't bots that are generating these conversations on HN. It's established users, who are divided on divisive topics. It's easy enough for us to ban accounts that are behaving like bots, and we certainly do. But the odds that a bot was posting about Julia in 2014 (I always use this example) are minuscule—no? > I honestly believe that it’s become more toxic and divided Yes, but because of the point I just made about established users, I believe this is explicable by divisions in the society at large. People (and especially internet users!) have a strong tendency to look for external factors here—blaming outside enemies, spies, shills, foreign agents, and so on—for our own problems. This is also a well-established historical pattern. I'm not saying it doesn't happen! I'm just saying we need to see specific evidence. If anyone has a link they think might lead to that, we're always willing to take a close look. But people should at least look at the public commenting history of such accounts before jumping to conclusions. I wrote a bunch about this here: https://news.ycombinator.com/item?id=27398725 https://news.ycombinator.com/item?id=27398725. If you think there's a problem with the argument there, I'd like to hear what it is. I'm not closed at all to the possibility that we're missing something and we need to do more. I just know what we've found after the last several hundred (if not thousand) requests for investigation that people have sent to us.
- crazydoggers 5y ago
  I think you may be underestimating 2 things, just from my previous experience. First, unless you are using sophisticated AI and behavioral data mining practices, you are not able to spot current bot activity. It is nearly undetectable from human activity by standard systems (ip address, browser fingerprinting, JavaScript systems, etc.). I know this because I worked in a space dealing with this. Google developed the tools to catch a lot of it, but does not disclose how they work. Their success at this is also why they now control most online advertising avenues. I know HN is small, and it may not be within your budgets for such systems, but perhaps there’s someone in the space willing to lend a hand? Second, a lot of this comes from non bot traffic, and regular user accounts. Even regular users on other services have admitted being paid by various “advertising agencies” to post or comment on certain topics. And I do get that HN is not the size of Facebook or Twitter, and so the issues are not the same. However, again, HN reaches some of the brightest minds in an important field of the US economy, so it’s not exactly an unimportant target. Anyways, if you see my comment history you’ll notice I’m extremely skeptical and science minded. I appreciate and applaud your need for hard evidence, which I can’t offer since I don’t have all the data. I do have experience with this stuff though, and on top of that, if you’ve gotten several thousand requests for recent investigation, I wouldn’t be so quick to chalk it all up to increased divisions in society. In fact it certainly is part of it, since that’s the whole point. It’s a feedback loop. Thanks for taking the time reading this and replying. I do appreciate it.

dang 5y ago

I have a bunch more to say about this but unfortunately I have to run out for a training. A couple super quick points though: > Facebook, Twitter, Reddit, but not HN HN is orders of magnitude smaller. Sites face qualitatively different problems at each order of magnitude. I'm not saying that proves anything, but it's a sensible reason why we don't see the data they do. > just a few thousands of bots generated spikes of conversations I can tell you from deep familiarity with HN's data that it isn't bots that are generating these conversations on HN. It's established users, who are divided on divisive topics. It's easy enough for us to ban accounts that are behaving like bots, and we certainly do. But the odds that a bot was posting about Julia in 2014 (I always use this example) are minuscule—no? > I honestly believe that it’s become more toxic and divided Yes, but because of the point I just made about established users, I believe this is explicable by divisions in the society at large. People (and especially internet users!) have a strong tendency to look for external factors here—blaming outside enemies, spies, shills, foreign agents, and so on—for our own problems. This is also a well-established historical pattern. I'm not saying it doesn't happen! I'm just saying we need to see specific evidence. If anyone has a link they think might lead to that, we're always willing to take a close look. But people should at least look at the public commenting history of such accounts before jumping to conclusions. I wrote a bunch about this here: https://news.ycombinator.com/item?id=27398725 https://news.ycombinator.com/item?id=27398725. If you think there's a problem with the argument there, I'd like to hear what it is. I'm not closed at all to the possibility that we're missing something and we need to do more. I just know what we've found after the last several hundred (if not thousand) requests for investigation that people have sent to us.

crazydoggers 5y ago
I think you may be underestimating 2 things, just from my previous experience. First, unless you are using sophisticated AI and behavioral data mining practices, you are not able to spot current bot activity. It is nearly undetectable from human activity by standard systems (ip address, browser fingerprinting, JavaScript systems, etc.). I know this because I worked in a space dealing with this. Google developed the tools to catch a lot of it, but does not disclose how they work. Their success at this is also why they now control most online advertising avenues. I know HN is small, and it may not be within your budgets for such systems, but perhaps there’s someone in the space willing to lend a hand? Second, a lot of this comes from non bot traffic, and regular user accounts. Even regular users on other services have admitted being paid by various “advertising agencies” to post or comment on certain topics. And I do get that HN is not the size of Facebook or Twitter, and so the issues are not the same. However, again, HN reaches some of the brightest minds in an important field of the US economy, so it’s not exactly an unimportant target. Anyways, if you see my comment history you’ll notice I’m extremely skeptical and science minded. I appreciate and applaud your need for hard evidence, which I can’t offer since I don’t have all the data. I do have experience with this stuff though, and on top of that, if you’ve gotten several thousand requests for recent investigation, I wouldn’t be so quick to chalk it all up to increased divisions in society. In fact it certainly is part of it, since that’s the whole point. It’s a feedback loop. Thanks for taking the time reading this and replying. I do appreciate it.