New here?
Create an account to submit posts, participate in discussions and chat with people.
Sign up
I've created a new version of the Obfuscator. What it does is this:

Ιt tаkеs lеttеrs frօm wrіttеn tеxt аnd cօnvеrts thеm іntօ еxօtіc chаrаctеrs thаt lօօk sіmіlаr. Αlɡօrіthms shօսld nօt bе аblе tօ rеаd іt, bսt hսmаns cаn. Thіs shօսld bуpаss thеіr cеnsօrshіp аlɡօrіthms, whіch еxіsts օn ΥօսTսbе, Twіttеr аnd cօ. Thіs pаrtіcսlаr pаrаɡrаph іs օbfսscаtеd fօr еxаmplе.

Αոᶁ ṭһіṣ օոе іṣ օbꬵսṣϲаṭеᶁ іո ṭһе һіɡһеṣṭ рօṣṣіbᶅе ṁօᶁе.

The spell checker cannot read any of the above (copy-paste it to test yourself). Version 1 wasn't too convincing, the comments got removed even though they were obfuscated. My suspicion was that the algorithms could read the text and censor comments despite that, but it is more reasonable to assume they were censored because nothing could be read. So it appeared like pure spam or like ASCII images. Perhaps I just overestimated them. Version 2 only selectively obfuscates entire words if they contain something of the following:

"nigg", "fag", "cunt", "idiot", "shit", "cuck", "simp", "fuck", "jew", "yenta", "coon", "nog", "gay", "homo", "suicide", "tranny", "troon", "jeet", "brown", "retard", "hitler", "nazi", "natsoc"

Other words remain intact, so that their algorithms are satisfied by having enough readable words. That means for example "nigger", "niggerish", "niggermode", "niggered", "niggerlover", "niggerball", "sandnigger" all get obfuscated. You should also be able to write about jews and Hitler unhindered, and call people faggots and trannies.

Don't go overboard and get reported too much - it could make them aware of this, enact countermeasures and possibly fuck up your account. I will try it for a while and tell how well it works.

Here: https://files.catbox.moe/vpnwlf.zip
You must log in or sign up to comment
4 comments:
TallestSkil on scored.co
7 hours ago 5 points (+0 / -0 / +5Score on mirror ) 1 child
>օbꬵսṣϲаṭеᶁ

That’s interesting. I haven’t seen a failed Unicode character for over 20 years. I wonder what’s missing on my machine that I can’t see your f replacement.

**EDIT:** Even stranger, [I *do* have it](https://i.postimg.cc/rppvn8c9/f.png). It’s just not rendering in this font for some reason.

**EDIT EDIT:** *Browser* problem. Works in Opera but not Safari, despite the OS knowing about the character. Weird.

What’s interesting about this is that it *should* work unless a site expressly restricts typing to the most basic ASCII set. It’s how spam e-mails bypass every filter in existence. Hell, even “legitimate” organizations like Publishers Clearing House will use “1OO” in their titles instead of zeroes because of that kind of filter.
PurestEvil on scored.co
6 hours ago 2 points (+0 / -0 / +2Score on mirror )
> unless a site expressly restricts typing to the most basic ASCII set

Well, YouTube for example does allow Chinese and Japanese characters, so they must operate on Unicode. Those emoji are also text characters depicted as images: 🐬
PurestEvil on scored.co
5 hours ago 4 points (+0 / -0 / +4Score on mirror ) 1 child
Good news, 1/1 is successful. My comment was this:

"The biggest scandal is that you called her a man. I have trаnnу fatigue." [reference to that Eliot foid and the movie with Helen of Detroit]

YouTube would normally shadow-hide the comment for "tranny", but someone liked it and responded even. That means it worked. Still, I'll have to see how it works for replies, because root comments seem to be censored less than replies.

I'm also expanding the blacklist. So if you have suggestions for words that are likely censored, tell me. I also added "negro", "kill", "dissi", "censor", "suicide" and removed "nog." It will be in the next update.
AndurilElessar on scored.co
4 hours ago 2 points (+0 / -0 / +2Score on mirror )
Great work and I admire your dedication to circumvent the constant online censorship... The problem with this is that even when it works it only takes one report, one learned catch from an algorithm, or an actual mod to see the comment and shadow ban it. Even after it's been getting upvotes and replies.

Channel owners can now input words & phrases into the algorithmic watchdog that will very quickly retroactively remove and instantly shadow ban anything referencing what they don't want to appear. I see this abused by AI slop channels constantly on YouTube. This channel moderation feature is also abused by outright shills, bad actors, and other assorted degenerate shitheads all over YouTube.

I think YouTube has given channel owners way too much control and censorship tools to use against the YouTube audience. It allows them to grow a following without ever experiencing criticism. Or their audience ever being made generally aware of anything shady they're guilty of. It's sinister.
Toast message