Half a year ago, I found a Facebook bug where if you post a comment that contains both a URL in the text and a picture as an attachment, it gets removed in a few seconds. About 24 hours later, you get a notification that the comment violates community standards for spam. I tried reporting the issue but it's hard to get an answer or progress.
I agree with what a couple of YouTube comments said:
> censorship like this, that simply tricks you into thinking you're alone or being ignored, is a form of psychological torment.
> This is extremely frustrating, I've experienced this countless times. Most of the time I don't even know what causes censoring. It's insane
> They censor the whole platform like that, comments just go straight into the black hole and you never know what words triggered it.
Imgur does this, too. Actually pretty smart from the platform's PoV. Just delete or "shadow" comments (or whatever you call it, when you can see your comments but no one else can) that contain certain words.
No need for moderation, people still use your garbage, while not being able to influence anyone else or affect your advertising.
when an algorithm censors one form of content over another, then it is indirectly supporting the alternative. which is tuned based on biases, those biases pad the pockets of like minded people within their social sphere, their circle of influence.
but the perception is there that this is what everyone wants, so a percentage of people (competition in many cases) gets censored.
if this was like a small corner bar and someone was advertising their business inside someone else’s bar, then yeah i get it, your house, your rules, but censoring someone on a platform that is larger than many nations put together seems dangerous.
But they're not being paid to do so, like this Law establishes. Unless they're doing it to bbenefit content that are more likely to generate ad revenue. But they already have the freedom to do so with their algos...
Maybe he could try to post another comment instead of the same one to pinpoint what's triggering this. There is a separate % sign, maybe a few words together causing filters to act.
His method is not proving anything because he keeps trying the same thing.
In the comments he describes the issue in more detail. In this case the word "poof" is being flagged for comment deletion and other people were encouraged to try posting it and their comments were also deleted.
I wanted to copy-paste the comment here but for some reason YouTube makes that difficult on my phone.
oh... that started much longer ago. Back in the earlier GUI days you could rely on copy/paste to work, now that everything is a web app, it's total pot luck if you can even copy/paste inside the same app, let alone across different apps.
This could be for multiple reasons - a cache consistency issue, the user could be blocked by that channel, or the comment could contain a channel blocked word.
I'd guess that if the automatic comment approval system isn't immediate then YouTube shows the comment to the poster so they can validate it was submitted and they haven't made typos but other users don't see it - and when the comment is checked (~20s later) and found bad it gets deleted.
Also I'm pretty sure that trying to post the same reply to the same user from a same account multiple times in a row will trigger some sort of spam detection.
The comment is still being processed by YouTubes comment checking system, the poster is probably the only person that can see the comment until it has been checked. As others have said its probably being removed for including the word "Poof"
So that can't be a yardstick -- there has to be some more objective criteria where a word is recognized as truly offensive by a large majority of the population.
I’m not sure that objective criteria will work here. Slang terms about sexuality often encode the bias of the majority and are targeted at humiliating a minority group.
Rather than trying for a universal measure, I think it’s best to leave it up to individuals, letting us all use whatever empathy and context we personally have.
It appears he was replying to someone else who had the same issue and was attempting to explain why that user’s comments were being deleted. I doubt those comments also contained the word “poof.” Also, the video in question contains numerous comments with the word “poof” that remain visible.
You said "channel block list". This word isn't on any such list. This was decided upon by YouTube and YouTube only, and is very clearly a deliberate one.
Not in defense of google, but public boards and places for comments illustrate that most posters are idiots who at best post irrelevant text or at worst… well many worse types of content.
What is particularly frustrating is that tons of garbage posts live on whole a few legitimate or at least not offensive posts get removed.
I have also noticed this when posting comments containing certain words (that seems to be what is happening here) or when posting links. It can be frustrating if you want to back up your comment with a source.
I guess it's understandable given YouTube's scale. Easier to just remove all comments with links than to do more advanced spam detection.
a number of the channels i follow have a regular set of porn bots that post links, and also there are also obvious (to a human) machine generated comments.
I assume (without any evidence) that YT has a programmatic / ML approach to attempt to minimise this sort of crap..
I note that in this case the problem seems to be a particular word, but i don't see this as malice on YTs part.
I no longer post links as references for answers etc even when the channel creator allows them, after having one YT account permabanned for spam based (as far as i can tell.,) on links posted...
Those algorithms are used to control discussions and censor things as well. This is completely obvious if you pay attention to what is removed.
The excuse is always to fight bots and spam, but it's exactly like how the war on terror is actually about war on human freedom - building a prison we all live inside.
Youtube has been using their platform to openly manipulate people's ideas since 2010 (it started as a tool to slow ISIS, but now you can get your comment deleted for mocking the CCP even in the US.) Use PeerTube, Odyssey, or Bitchute instead.
If I remember correctly people demonstrated similar things in 2018 or 2019 where you could see your comment but if you lived or it was gone, kind of demonstrating that only you could see it.
Assuming it is the word 'poof', a term, at least in my circles, which hasn't been in current use for at least thirty years I'm intrigued to know how YT deals with obscure slang in, a language with many fewer speakers.
Are mildly derogatory terms in Burmese (1/10th the speakers of English) treated in this way ? I'd be curious to know how an Anglo-centric enterprise such as YT can establish what semi-archaic derogatory terms there are in Burmese. I suspect the comment would be safe from deletion just as long as English speakers can't read it.
> someone posts something concerning on hn about cultural issues or govt or tech censorship
Autists on HN: "I m highly interested in some stupid unrelated thing that brings no light to the main topic which is big tech censorship or biased moderation, but diverts attention of the users ultimately producing no valueable exchange of ideas."
It s not just you btw. There are many users that do exactly that on any hot topic on hn, starting entire threads that span dozen of comments on irrelevant things. Not sure if it s a cope, or a "bury your head in the sand" psychological thing, or just a trait of being in tech.
OK, good point, I'm going to plead guilty. I didn't set out to do that and I am concerned about big tech censorship and the like but nevertheless I accept that was the effect. Mea culpa.
Look. It s not a personal observation about you. You can obviously do whatever you want and comment whatever you please. It is just an observation about the way many users on hn react, that s it. No need for a mea culpa.
Even if it's archaic it's still offensive and people will use it to abuse. The problem is that YT isn't able to determine context, which is surely a much harder problem than word match. I imagine.
YT does any number of things in multiple languages. Offensive words in Burmese or any other language can easily be implemented with just a word list, which shouldn't be that hard to derive, for example you could scan a dictionary that uses the 'offensive' tag on words like English dictionaries do.
I recently tried to post a comment on YouTube in a video where someone used Blender for modelling some icospheres for woodworking and I realized that there is a bug that produces irregular geometry.
I don't know why but it was silently deleted. Maybe because I don't have my real name on YouTube, maybe because it mentioned filenames of the source or whatever or maybe I'm just not paying google enough for them to publish my comments.
There's a youtube video where someone suggests deliberately brewing with moldy sloes. This is dangerous because of mycotoxins that drop from the mold into the brew and cannot be removed by boiling or any other technique available to the home brewer. (I won't link to the video to avoid giving it google karma, but it's by "brewbitz").
My comment that this was dangerous was deleted by the channel owner and apparently I was blocked. So the video suggesting people do a dangerous thing lives on.
Had this happen to me. At first I assumed its because I block most of Google/YT tracking (pagead, things like https://www.youtube.com/youtubei/v1/log_event?xxxx), but unblocking everything didnt fix the issue. It would sometimes not delete a comment if I manually reposted it a minute later.
I very rarely post comments on youtube, but when I do, I often want to go back to see if I've gotten any replies. Anyone else who's tried doing this has figured out that youtube makes it incredibly difficult for you to find your comment history. It's clearly not something they intend for you to actually use. They don't give you notifications about replies, or even the option to be notified.
Anyway, I think a lot of my comments have either been silently deleted or that maybe they work as if I'm shadowbanned. I've seen this in action trying to post comments on a news video.
It must have something to do with long comments. Whenever I make a long comment, most of the time it just gets deleted after I reload the page. I even checked out the comment history which also shows the comment at first, then you reload and it disappears.
Then after 10 retries, I say fuck it and I just write a small comment instead, go back after 20 seconds and it's still there.
It probably has something to do with long comments. Not sure, but that's where the clues are leading me.
YT Bell icon will show you replies to your comments, but only those where someone directly clicked "reply" on your comment. YT comments are 1 deep and if someone replies to a reply to your comment you wont get notified.
But this wont show deleted comments. I had comments disappear, and later video creator tell me he saw it (got notification) before it was automagically deleted by Google, those comments vanished from YT history.
Google and all the rest give users the impression that the comments section, or in the case of social networks the feed, is a sort of digital market square of organic conversations happening unhindered between participants. It is not and never was. It is only recent events of the last few years that have made this abundantly clear.
similar phenomenon for different reasons on Reddit: anytime I felt compelled to write up a comment, within minutes I would receive a notification that my comment was removed since I wasn’t a regular member of the community. How to become a regular when I’m not allowed to comment is a mystery to me, without even getting into how every subreddit has a different auto-moderator working off some opaque spam/profanity filter.
how much content is produced but never published online, silenced by robots?
76 comments
[ 3.0 ms ] story [ 136 ms ] threadI agree with what a couple of YouTube comments said:
> censorship like this, that simply tricks you into thinking you're alone or being ignored, is a form of psychological torment.
> This is extremely frustrating, I've experienced this countless times. Most of the time I don't even know what causes censoring. It's insane
> They censor the whole platform like that, comments just go straight into the black hole and you never know what words triggered it.
No need for moderation, people still use your garbage, while not being able to influence anyone else or affect your advertising.
https://en.m.wikipedia.org/wiki/Payola
but indirectly
when an algorithm censors one form of content over another, then it is indirectly supporting the alternative. which is tuned based on biases, those biases pad the pockets of like minded people within their social sphere, their circle of influence.
but the perception is there that this is what everyone wants, so a percentage of people (competition in many cases) gets censored.
if this was like a small corner bar and someone was advertising their business inside someone else’s bar, then yeah i get it, your house, your rules, but censoring someone on a platform that is larger than many nations put together seems dangerous.
We don't shadowban established accounts—we tell people we're banning them and why
:
The exceptions are spammers and serial trolls.
Some previous notes:
https://hn.algolia.com/?query=by%3Adang%20shadowban&sort=byD...
His method is not proving anything because he keeps trying the same thing.
I wanted to copy-paste the comment here but for some reason YouTube makes that difficult on my phone.
Is that a millennial dog-whistle for something unsavory?
Using a platform to complain about said platform is due to lack of choices, not hypocrisy.
But you're also right, if people don't complain and try to improve something, the chance of any change is much lower.
https://creatoracademy.youtube.com/page/lesson/connect-with-...
I'd guess that if the automatic comment approval system isn't immediate then YouTube shows the comment to the poster so they can validate it was submitted and they haven't made typos but other users don't see it - and when the comment is checked (~20s later) and found bad it gets deleted.
A perfectly normal word...
Even “gay” which is popularized and usable in casual conversation is still considered offensive by some people because of its historical aspect.
For example read up on urban dictionary for some random word -- say omelette:
"Slang word for a sexual orgy." https://www.urbandictionary.com/define.php?term=omelette
So that can't be a yardstick -- there has to be some more objective criteria where a word is recognized as truly offensive by a large majority of the population.
Rather than trying for a universal measure, I think it’s best to leave it up to individuals, letting us all use whatever empathy and context we personally have.
I don't want to argue about the specific content, but Google made it very clear they just remove content they don't like.
So they admit they censor content. And because it is their platform you can't do anything about it.
What is particularly frustrating is that tons of garbage posts live on whole a few legitimate or at least not offensive posts get removed.
I guess it's understandable given YouTube's scale. Easier to just remove all comments with links than to do more advanced spam detection.
I assume (without any evidence) that YT has a programmatic / ML approach to attempt to minimise this sort of crap..
I note that in this case the problem seems to be a particular word, but i don't see this as malice on YTs part.
I no longer post links as references for answers etc even when the channel creator allows them, after having one YT account permabanned for spam based (as far as i can tell.,) on links posted...
(anecdata is notoriously lacking in peer review)
The excuse is always to fight bots and spam, but it's exactly like how the war on terror is actually about war on human freedom - building a prison we all live inside.
No, it is not. Without these algorithms Youtube will be flooded with comments from bots.
> Those algorithms are used to control discussions and censor things as well. This is completely obvious if you pay attention to what is removed.
May be I'm not paying attention but this is not obvious to me. Care to share evidence for your claims?
You never actually see censorship if you are not censored.
If he/she "doesn't see it because he's not beingu censored", as you seem to imply that perhaps you are, his question would then be very relevant:
You could enlighten them with evidence so that they can also see what is obvious to you.
Are mildly derogatory terms in Burmese (1/10th the speakers of English) treated in this way ? I'd be curious to know how an Anglo-centric enterprise such as YT can establish what semi-archaic derogatory terms there are in Burmese. I suspect the comment would be safe from deletion just as long as English speakers can't read it.
Autists on HN: "I m highly interested in some stupid unrelated thing that brings no light to the main topic which is big tech censorship or biased moderation, but diverts attention of the users ultimately producing no valueable exchange of ideas."
It s not just you btw. There are many users that do exactly that on any hot topic on hn, starting entire threads that span dozen of comments on irrelevant things. Not sure if it s a cope, or a "bury your head in the sand" psychological thing, or just a trait of being in tech.
I do appreciate all efforts to keep the conversation polite though so have an up-vote ;-)
YT does any number of things in multiple languages. Offensive words in Burmese or any other language can easily be implemented with just a word list, which shouldn't be that hard to derive, for example you could scan a dictionary that uses the 'offensive' tag on words like English dictionaries do.
I don't know why but it was silently deleted. Maybe because I don't have my real name on YouTube, maybe because it mentioned filenames of the source or whatever or maybe I'm just not paying google enough for them to publish my comments.
We need to replace them now, before it is too late.
My comment that this was dangerous was deleted by the channel owner and apparently I was blocked. So the video suggesting people do a dangerous thing lives on.
I just gave up and stopped commenting on YT.
I very rarely post comments on youtube, but when I do, I often want to go back to see if I've gotten any replies. Anyone else who's tried doing this has figured out that youtube makes it incredibly difficult for you to find your comment history. It's clearly not something they intend for you to actually use. They don't give you notifications about replies, or even the option to be notified.
Anyway, I think a lot of my comments have either been silently deleted or that maybe they work as if I'm shadowbanned. I've seen this in action trying to post comments on a news video.
It must have something to do with long comments. Whenever I make a long comment, most of the time it just gets deleted after I reload the page. I even checked out the comment history which also shows the comment at first, then you reload and it disappears.
Then after 10 retries, I say fuck it and I just write a small comment instead, go back after 20 seconds and it's still there.
It probably has something to do with long comments. Not sure, but that's where the clues are leading me.
also https://www.youtube.com/feed/history click Comments on the lower right.
But this wont show deleted comments. I had comments disappear, and later video creator tell me he saw it (got notification) before it was automagically deleted by Google, those comments vanished from YT history.
how much content is produced but never published online, silenced by robots?