i saw tiktoks online before where people censor words in text in videos, assuming i'm correct doesn't that mean that there's an AI that reads this text?, hence it's censored so the ai won't shadowban the user?.. by that logic, can't you write malicious code that can fuck up the ai, and once the ai reads it it gets brainfucked?
(to note, i don't watch tiktoks.. i just saw some before.)
> I have no idea what I'm talking about, the post
well, you're not wrong at all.. it was just an assumption i don't really understand how this thing works lol
>i saw tiktoks online before where people censor words in text in videos
that's just mutt hypocrisy
>doesn't that mean that there's an AI that reads this text?
stop drinking the ~~*AI*~~ koolaid, if there is something it's braindead software like the rest, not that I'd know as I avoid vertical videos altogether
>by that logic, can't you write malicious code that can fuck up the ai, and once the ai reads it it gets brainfucked?
what you see in hollowood propaganda does not translate to real life
>it was just an assumption
you're projecting way too much, you obviously don't know shit about technology, know your fucking place and sefl-censor yourself before posting this kind of garbage. you know you're not qualified to have a correct opinion, what makes you think you can articulate a question that deserves an answer?
taht's not how software work, you giving too much credit to ~~*AI*~~
holy shit you're a gay
YouTube literally has a bot that detects anything remotely resembling guns, mentions of violence and terrorism, and explicit use of copyrighted materials. It's not far-fetched to assume that it works, but you'd need to have insider information on how their site was programmed in order to be able to exploit anything sussy with it
>YouTube literally has a bot that detects anything remotely resembling guns, mentions of violence and terrorism, and explicit use of copyrighted materials
so what? this is about tiktok and mutts are known to do that all the fucking time for no reason, the very few tiktok videos that I've seen were about gun violence in the us, nothing was censored so I'd bet on dumb self-censorship.
>It's not far-fetched to assume that it works, but you'd need to have insider information on how their site was programmed in order to be able to exploit anything sussy with it
does not change the fact that you can't somehow "brainfuck" a fucking parser just by feeding it hand-crafted payloads...
if the software does not detect bad stuff then it will process the next videos, what's so different between gibberish and legit content? both have no bad words in them so it will just work...
you can't "trick" the software, it's not intelligent, it's not human, it just read inputs and send an output, unless you know bugs exist in the software they use it's not realistical doable to crash their system.
feeding random inputs to ML software is the corner stone of this field, don't think for a second that you'll manage to "brainfuck" anything
You're clearly either a BOT larper pumping out tech jargon buzzwords or a web dev so your opinion really doesn't matter regarding cybersecurity, but just so that misinformation doesn't spread I'm going to post this basic tutorial for anyone interested in knowing how to actually exploit a parser
?t=384
didn't watch
OCR isn't parsing, NLP isn't parsing and if you take malformed input into account it's very easy to protect your parser
>N-no... it's literally impossible for you to brainfuck a program, those kinds of programs don't even exist
>Uhhh... okay, they do exist but not in Tiktok because... they just don't, okay??? You still can't brainfuck a program...
>O-okay well you can but that doesn't count!!!! It only counts when I say it counts!!!! Besides, you can protect against those kinds of attacks easily
You know you can just admit you're wrong, right?
this is BOT, you should be happy with what you already got.
besides like it matters, it's some anonymous shmuck who watched two ltt videos and now thinks he is an it genius.
i'm nta, that was my first post in this thread
i only replied because you were spreading misinformation
>It's not far-fetched to assume that it works
it is
>You're clearly either a BOT larper pumping out tech jargon buzzwords or a web dev
firmware swe
>really doesn't matter regarding cybersecurity
I harden firmwares and linux for a living
>but just so that misinformation doesn't spread I'm going to post this basic tutorial for anyone interested in knowing how to actually exploit a parser
I should not have reduced ML inference to a parser, my bad, I hsould have known that autists would be stuck on the word forever and no discussion would be possible...
OBVIOUS FUCKING LY you can crash software, who would have thought...
my point was that you can't crash such tool if you feed it a correct video but with somehow hypnotizing data that would "brainfuck" it, this makes no sense and feeding wrong data you some NN is normal and expected.
>can't i ask a question without having a condescending rotten sweat reeking incel lash out on me
why ask a question when the answer is widely available on the internet? (for the censorship part) that just tell people that you're moron and you're not worht answering as you'll forget about it a few nanosecond after you closed the thread.
retards like you should be discouraged to even interact with other people
> because i know less then his superior knowledge about everything?,
I don't know shit either, I search everywhere I can BEFORE asking retarded questions also there is a thread on BOT for stupid shit where retards like you can ask such dumbshit question
yes tiktok censors text, I just checked google and got the answer under 10 second
no you can't "brainfuck" software
delete your shit thread and post my incel answers on your favorite tumblr page or subreddit
>gets upset that people are stuck over the fact that he said parser
>is stuck over the fact that OP said brainfuck
damn, BOT is well known for it's very.. well, "unpleasant" user base, but you my friend.. take the cake, one of the biggest assholes i saw today, can't i ask a question without having a condescending rotten sweat reeking incel lash out on me because i know less then his superior knowledge about everything?, you're pathetic
holy shit you know nothing about IT
uh.. btw, why exactly wouldn't it work?
can't i, make a BOT thread without having people arguing in the replies?.. instead of being so arrogant and full of yourselves, just because you are anonymous doesn't mean you can't have a bit of decency...
the guy being a cunt is a cunt but he's right. you can't "brainfuck" or crash an A.I. using NLP / ML. it's a category of programs where "wrong" input is expected at all times. ML especially is very hard to crash outright, though sometimes you can make it give unexpected outputs.
in real life the famous "DROP TABLE shit;" jokes don't actually make sense, writing code in text boxes is almost never an attack vector.
sorry op your thread got immediately ruined by a schizo