Hacker News new | past | comments | ask | show | jobs | submit login

>It looks like Twitter is suppressing posts until they are spammed by hate bots and then making those posts visible.

>https://bsky.app/profile/willhaycardiff.bsky.social/post/3lk...

This could also very well be explained by a ranking algorithm that optimizes for "engagement". Getting spammed by hate bots = "engagement". This would be perfectly consistent with what the guy is experiencing, minus the accusation that the platform is suppressing anti-ukraine posts, which is totally unsubstantiated.




As I understanding the timing, the post was suppressed until the hate bots spammed it.

Given the post was suppressed, how did the hate bots know about it to spam it?

It seems to me Twitter suppressed the post until they had time to spam it with hate posts.

Bear in mind here also this suppression did not happen for other posts - only for the pro-Ukraine post - so Twitter at the least is specifically suppressing pro-Ukraine posts.


>It seems to me Twitter suppressed the post until they had time to spam it with hate posts.

Twitter can't spam hate posts in real time? They're literally an AI company.

>Bear in mind here also this suppression did not happen for other posts - only for the pro-Ukraine post - so Twitter at the least is specifically suppressing pro-Ukraine posts.

There's scant evidence of this, besides the vague assertion that "I have 5k more followers there and regularly have posts which are viewed over 100k times". If he normally posts informative and substantive content that gets good engagement, is it really surprising that a generic "I support Zelensky" post would get poor engagement?


I’d bet that their tech is weaker than what people expect. Perhaps a better term would be “not as robust as expected”.


> Twitter can't spam hate posts in real time? They're literally an AI company.

I would expect they have a certain capacity and the flow of posts is very, very, very great.

Also, I would imagine a delay of some sort anyway - people would notice if a post immediately received hundred of one line generic "I hate this" replies. In this case, the hate posts began after 12 hours.

> is it really surprising that a generic "I support Zelensky" post would get poor engagement?

I may be wrong, but as I understand it, what you've described here is not what was described by the author.

As I understand it : one guy, two accounts, both long term, one BSky, one Twitter, same content. Approx same and large number of followers. On BSky, the test post has immediate, large-scale and positive response. On Twitter, this and this only post suppressed for IIRC 12 hours, then a massive number of hate-bot posts and then post made visible.




Consider applying for YC's Fall 2025 batch! Applications are open till Aug 4

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: