Huawei 5G kit must be removed from UK by 2027

Internet users are a thousand (actually probably more like a hundred thousand) times too quick to jump to such insidious but exciting conclusions. Having spent countless hours investigating such things I can tell you confidently that the overwhelming explanation is the boring and obvious one, the one Mr. Occam will give if you ask: people just disagree.

People are biased toward underestimating how much legitimate disagreement there is in any large, distributed population sample—which HN is. Probably we're hard-wired to evaluate the world by local conditions around us, and most of us are surrounded by people who see things similarly to how we do. Then we come online, bump into views that are harsh outliers in our world, and poof: an astroturfer under every bed and a spy in every closet.

This is so close to a universal mechanism that we have a rule about it in https://news.ycombinator.com/newsguidelines.html:

"Please don't post insinuations about astroturfing, shilling, brigading, foreign agents and the like. It degrades discussion and is usually mistaken. If you're worried about abuse, email hn@ycombinator.com and we'll look at the data."

Please note: this is not to say that abuse doesn't exist. But the overwhelming majority of such insinuations are imaginary, so in investigating real abuse we need concrete evidence to go on—something, anything. The presence of opposing views on an internet forum does not clear that bar—it is evidence of nothing but that the topic is divisive.

Piles of past explanation are at https://hn.algolia.com/?sort=byDate&dateRange=all&type=comme...

I wonder if you can talk a little about how you determine whether or not astroturfing is going on. How can you be sure that a given post with a political or nationalist slant is made in good faith? You seem pretty confident that it's not a widespread problem, which is great, but it would be good if the rest of us could share at least some of that confidence.

Is it really enough to look at the account history and observe that the same user has participated in other unrelated threads? Or are there more subtle cues that you go by?

Does HN have anything like Twitter or Facebook to detect organized network of sock puppets?

I don't know what they do so I can't say. We certainly have lots of code dedicated to such things.

We can't know for sure. Past account histories count for a lot—if people would simply look at them, I think most accusations of this sort would vaporize. It's not plausible that $secret-enemy planted $hn-user in 2013 to comment about syntactic whitespace and Google Fiber or whatever, so they'd be more credible seven years later when promoting $political-agenda.

Beyond that, we look at relationships between accounts, patterns of site access...I'm not sure what else to tell you. The private data confirms what the public data already shows. There are exceptions, but they don't determine the discourse on the site. What determines that is people simply having different views.

Mostly I just wish people would realize that the spectrum of genuine disagreement is much broader than it seems like it should be, would be, or could be. The world is just a bigger place than we feel like it is.

That brings a surprisingly amount of false positives. After all nobody would think that two users might share a proxy ip address.

We're pretty familiar with users sharing IP addresses.

You're right that software defenses inevitably produce false positives, so have to be used carefully.