[grecy]

Thanks for proving my point so succinctly!

For the record I'm not complaining about "genuine" Chinese people expressing their opinions. As you said, it's important they have a voice, and we should listen.

What I'm pointing out is the increasing number of accounts that are very, very obviously paid or otherwise government controlled to influence opinions on HN (and obviously elsewhere on the internet)

[dmix]

Do you have some examples? It'd be good if we could keep a list and screenshots of this sort of thing, with some research into the usernames.

I hear people on Reddit talk about 'Russian bots' constantly and I've always wanted to see some examples on a site like HN/Reddit.

The above person lists his Github account and AFAIK he's not a paid shill, just a political contrarian or provocateur for political ends. Which IMO is an important difference if we're going to accuse everyone of being bots and "paid shills".

A green name with a single comment being downvoted immediately isn't influencing opinions here, they can't even downvote. But it'd be interesting to measure their frequency as well for a research project.

[xfs]

At the risk of being labeled one myself... there are no examples. I've seen dang respond to complaints like this many times and every time it was real people with real profiles being accused of shills, trolls etc.

Also, if you look it up it's almost all politics threads wrt China this year on HN with overwhelming anti-China sentiment, yet here we are, arguing about pro-China influence being too strong.

Then the question is no longer about how many shills are objectively out there that can be quantifiably measured, because this fact is obviously not what people have been basing their accusations on, but what results in the very question of shills being raised in the first place.

The only theory I can think of would be Chomsky's fifth filter, that is, a common external enemy that helps maintain consensus and divert ideological stress from internal antagonism, be it terrorists, Russian trolls, Chinese shills. This is compounded by the universalist belief that it is impossible to hold "genuine" political thoughts other than the End of History liberal democracy project, which is in itself beset on all sides already, making it all more intense.

[loceng]

The vast majority of users doing this will be real people who're simply indoctrinated - it's why China's tyrant leadership and censorship/control mechanisms is so effective, albeit powerful - however I don't like wrongly conflating the idea of what true power is with control.

[grecy]

So in this discussion [1] Checkout all the replies from 'baybal2'. Like this one [2].

Now look at the submissions for that account [3].

Now read more of the comments [4].

That account has been around a while, so I'm not sure it qualifies as "shill", but it certainly seems suspect to me.

[1] https://news.ycombinator.com/item?id=21124115

[2] https://news.ycombinator.com/item?id=21126141

[3] https://news.ycombinator.com/submitted?id=baybal2

[4] https://news.ycombinator.com/threads?id=baybal2

[dang]

You couldn't be more wrong, and the example shows how rooted these discussions are in imagination and skirt the margins of some truly ugly behaviors, which I'm certain you would never knowingly engage in.

I know who baybal2 is—we've exchanged perhaps a couple dozen emails over several years. I know his name and nationality. (Unless you want to argue that he's been emailing under a false identity? That's what spies do, after all.) He's someone with a technical background who's done extensive business in China. His views come from those experiences and no doubt from the rest of his background. This gives him a perspective that's very different from that of more mainstream HN demographics. Do we want a community member like that here? Or would we prefer to hound him out with suspicion and insinuation? Of course we want a community member like that here.

Why all the emails? Because for a while we were repeatedly banning and/or penalizing his account when it broke the site guidelines. When we think a user is persuadable, we'll often try to persuade them by email to use HN in the intended spirit. baybal2 may not have fully cleared that bar, but he's come a long way and that counts for a lot. And if you read his emails you'd see that he's a nice guy who means well and mostly has no idea when he's breaking the rules here; in other words, much like you and me.

Had you taken the time to really look, you'd find posts about traveling wave reactors (https://news.ycombinator.com/item?id=21175545) and biaxial helicopters (https://news.ycombinator.com/item?id=21168582) and Economist articles (https://news.ycombinator.com/item?id=20950878) just in the last month. This is not how "suspect" accounts behave. This is how normal users, who are as intellectually curious as you and simply have a different background and perspective, post to HN.

In most cases, you can easily figure this out simply by taking the time to look at an account's public posting history. Unfortunately, what people seem to do instead is see a handful of data points—and when I say "handful" I'm being generous—that pattern-match a pre-image they have in their minds ("pro-Chinese agent" or whatever). From those few data points, they autocomplete the rest of the dots into a sinister picture—the picture they already had to begin with. Once they've done that several times, a feeling of pressure builds up that they call "overwhelming evidence" or something like that, which they can't help but vent into the threads. This is the real problem, not the posting history of someone like baybal2.

I feel ambivalent about writing this. On the one hand, it's important to look at specific examples that illuminate how this internet phenomenon of accusing others of astroturfing, etc., fundamentally comes from projection: reading into external situations the image that one carries in oneself. This community badly, deeply needs to take that insight in.

On the other hand, it feels sickening to pick apart individual histories in public. Because we have baybal2's email address, I can at least check in with him. But there have been other cases where that wasn't an option, including this one: https://news.ycombinator.com/item?id=19403358, which had a mixed outcome. The user who made the accusation responded magnanimously. Unfortunately, though, the accused user really was hounded off HN and never came back. IIRC, they sent an eloquent email but refused our invitation to keep participating—or maybe that was someone else. There have been many such cases, including one that's sitting in the inbox right now, that I have yet to figure out how to reply to.

Is that really the community we want to be?

[dmix]

Thank you dang. I often think about how great of a job you're doing here and wondered if I could build a Reddit replacement built on your approach to community building. It's such a challenging problem to solve. I wonder if a strong creed and hiring approach could scale up to a larger site but I have my doubts.

It can't be easy doing your job and I hope you understand we very much value your work, even if everyone doesn't realize it's being done.

[baybal2]

Hello,

What would make you to reconsider? I don't really hide my identity on the Internet, and see no reason to do so for as long as I want to have a life.

Can you be one of those men who trolled me and my coworkers on email in July? If so, you need to work harder "to break my life." We were having good laugh reading that silly correspondence on lunch brakes. Very glad that I work in China now, and that people here don't give a f* about such drama, unlike in US.

[inetknght]

Amateur spammers will create new accounts as needed.

Well-run psyops will build and curate accounts over a longer period of time.

[loceng]

What I want to see is the HN data - anonymized if necessary, so we can calculate averages - and know what accounts are behaving in what brigading behaviour; and so in essence the community of HN can crowdsource moderation - at minimum to help spot and highlight patterns, so each reader or user can independently analyze, interpret and decide for themselves what the data means.

[grecy]

Right. I'd love to see something like this done for HN.

https://twitter.com/AirMovingDevice/status/11811206016430735...