by Taryll02
Since 2020 iv noticed some comments disappearing
And then soon after a drop off of responses I would get from comments
Usually polite and political stuff
But nothing malicious
I like drama/conflict but in a debate sorta way
I like clashing ideas not following a mantra
Sometimes I would swear being Australian so I assumed a similar but more passive chat filtered you see in games
I was a little annoyed by this cos it was unclear what words would trigger it
I did some research and nothing like this was being announced
I didn’t think much but kept using YouTube over the following 2 years
The filter seems the be ever changing
A year ago it would be cliches like, don’t combine “bad word” with “opressed group” I.e Women, Gays, Trans, Feminists e.c.t
This made sense as since the CEO become a women
And that it would be biased in that way, mainly because no one else is complaining to remove words and ideas as much as them
They used to be fractured as individual groups, but now they’ve banded together under the shadow of companies that suck their attention and money, similar to political voting
But as of late 2021 iv noticed it changes behaviour after viewing it for too long
It would normally delete exactly 17-19 seconds after posting (suggesting a list)
But after a month it would vary
It would delete obvious ones in that time frame
And later delete more complex wording 35-80 seconds later
Seems random, or bottle necked by something, most likely processing the comment
I think also to preserve processing power they catagorise accounts by the history, both videos and prior comments
I wanna say they can apply personality types and temperaments but that’s a stretch
With this information they can, hide certain content from certain types of people
And provoke certain reactions by promoting certain content to certain people
I believe this applies to comments aswell
They could have many channels that intertwine without conflict
Allowing certain types of people to mingle and repel others
This could be enterpretted as them trying to prevent conflict
But I doubt that
Iv also noticed that the access to a coherent and live (which is completely and was possible) comment history
Even with it all angled and hidden through google searches
They rearranged again to bury the feature deeper within the site
Why put all this effort to rename things and change the ui
Just for it to not gain any benefit but actually lose out
Maybe there is a benefit but not for us
This would make observing the AI live almost impossible
And can also use your cross platform activity to find those who could access detrimental information
This is a stretch but I feel it’s true as I would started recording any interaction with the comment section
I would test terms and watch it change to even accomodate men and “not opressed” groups
But it was personalised for my device/ip/account
Other people with a less provoking track record could get passed issuing these terms but would update later for them
Often a minute or two later
It allows you to make “bad” comments
But learns you are a “bad” person and prioritised you
I think as a response to this or people noticing they’ve also made it so other accounts can’t see your comment but it remains in your history
But established bannable word combinations are instantly banned like “what’s the point of hiding those”
It’s like they have no intention of informing us so we can learn
But to remove them as the natural responce before it’s the status quo
This effectively hides ideas that might be rampant in EVERYONE but makes everyone feel alone to kill ideas or social movements
Not allowing Monopolies is companies was to prevent blackmailing/forced markets/options
This isn’t applied to social media’s
or most internet based companies
This forces ever piece of information through them