Blackbluue

Just got my new model, hopefully I'll be able to show it off on stream soon. It's been a long time coming 🎨: @maddnesswriter #Vtuber #VTuberUprising #ENVtuber

@GeneFirewind @tallcowyt what third party are you referring to?

@BlackbluueVT @tallcowyt 1) Discord doesn't run the system, it's a third party 2) It absolutely bans people and does so with large error margins I've collected enough data on false bans over the last year to see it then doesn't allow the tickets to be seen by mod staff, seemingly to cover its tracks.

Sooo since Discord Support responded to my original post, I would like to respond to their response: 1) A Discord employee replied to my original post saying that it was “misinformation” and that “this has nothing to do with AI” which, no offence to them, but I think this is nonsense. For example, this official Discord article: discord.com/safety/how-dis… explains how Discord uses “machine learning”, which is a subset of Artificial Intelligence. Also, the employee misunderstood my post, I was trying to make a general point about the problems with AI moderation on Discord and they took it as if I was only referring to that one specific issue. (No hate to them obviously) 2) Adding on from that, I do believe that if Discord has to use AI for moderation, it has to be changed. The current system is fucked up and causes thousands of people to falsely lose their accounts when they have done nothing wrong. The incident this weekend was one example, but not just that, MILLIONS of people get false bans because of AI all the time!! This even happened to an owner of the unofficial Roblox server, a server which has 1m+ members and is a partnered server with Discord themselves! Link: x.com/notexttospeech… The best way to fix this going forward would be to, at the very least, give everyone an option to choose to have a human reviewer for all support tickets & appeals. I’m pretty sure Roblox already does this but I could be wrong, regardless of that it would be great if Discord and other companies like YouTube adopt this. A lot of the time, clankers are stupid and make unfair decisions, which can end up ruining people’s lives. A lot of people rely on their accounts for stuff like work or to communicate with friends and that can be taken away from them because of something that isn’t even their fault. 3) Discord claims that they’ve unbanned everyone, however I’ve seen a lot of people who are still banned because of sending grid images. Yeah… soooo… they can’t just be lying to us like that. Feel free to like & retweet to spread awareness and drop me a follow if you want. Also Discord please increase the max friends limit. 1000 is too low, it would be nice if it was 2000. And let us record audio in VCs and stages on mobile and iPhones. Thank you and have a great day!! @discord @discord_support @notexttospeech

@easkateconcepts just curious, but how were you planning on proving your identity to turn off 2FA? (don't actually post the stuff, just asking *how* you would do it)

update: finally reached a human (it took a tweet with 15k views) discord’s official answer after wrongfully locking me out is that I should just ‘create a new account’ I have no backup codes, so a decade-old account, years of important contacts is just gone, and the fix they offer is ‘start over’ man.

@vo5100 @PepperFyves whether AI moderation is a problem or not is besides the point. this incident has absolutely nothing to do with AI, so claiming that it does is detracting from the conversation.

@BlackbluueVT @PepperFyves Yes but they attached that to an auto-ban bot system & understaffed their support staff in favour of ai moderation tools... That's the problem

Fortunately I was not flagged by their system but I could have easily been. In the past I had done work that involved tartan patterns and I'm currently working on a flag design. This is why you don't replace all of your support staff with AI. You've lost a lot of trust.

@ScrubaatiVidya what you are arguing is essentially to change gun laws because someone got their purse stolen at knife point. terminology 100% does matter when you are claiming something that has absolutely nothing to do with the situation.

@BlackbluueVT No one, literally no one gives this much of a fuck about “correct terminology” because everyone knows what’s being talked about and being “correct” is irrelevant, doesnt change in the slightest they let a machine permanently ban people based on poorly coded crap

When will discord understand letting AI handle bans on that leave is just not good BECAUSE of this? False positives are why I’m still convinced my account got false banned over something I’m still sure was not csam

@RvLeshrac @dootskyre @panley01 do you need an AI to tell you 1 = 1? because that's what you are claiming right now

@BlackbluueVT @dootskyre @panley01 Every incident that involves detection of images involves MLAI.

To the people screaming "AI!!", this is something worse. Discord uses PhotoDNA fuzzy hashing to identify CSAM. One innocuous photo, one innocent frame in a video, that looks just enough like your upload? Bam. PhotoDNA is the industry standard & NCMEC endorsed. But... 🧵

@ScrubaatiVidya you do understand that "automation" is a term that is not directly tied to AI right? we have been automating systems long before AI was ever even a concept.

@BlackbluueVT What the fuck do you think is automating their moderation? A fucking hamster on a wheel and a fish in a tank? The fucking stock market chicken from south park?

@ReimansZuwa @ScrubaatiVidya 2 approaches to moderation with vastly different benefits and draw backs. you can't lump them both together under any umbrella term and say "it's all bad". because there is no viable alternative at this scale

@ReimansZuwa @ScrubaatiVidya automated systems are fully planned out by the developer. there is no ambiguity; putting a 1 into the equation always spits out a 2. we know exactly what it will do. conversely this limited and restricted subset of outcomes means it isn't malleable to unforeseen circumstances.

@Feyfolkenn @Paul__Walsh you should still file the appeal. the number is low, but there are people who have been granted appeals. the alternative is just giving up and not using discord at all.

@BlackbluueVT @Paul__Walsh Their support tickets seem to all be reviewed by the same sort of AI that mistakenly banned the users in the first place. You can't contact humans through their support portal, so there's no point.

Discord once banned users by falsely flagging grid images as CSAM. Now they want real identity verification. Future false bans will mean permanent lifetime bans since new accounts will be impossible to create. This team desperately lacks real tech expertise.

@BiggerDinger @SavageWuff discord can't do anything about it. they have no control over PhotoDNA. Microsoft has to fix it.

@BlackbluueVT @SavageWuff ur right but i dont think discord will do anything about it unless a post pops on the timeline with 10k+ likes cause they are lazy

Discord core thread🧵: >Devs lying about exploits not existing. >Devs lying about human mods reviewing CSAM bans before applying them. >lied about removing those bans ( grid/spiral ) exploit ones. >exploits still exist even though i reported them 5 months ago including grid one.

@RvLeshrac @dootskyre @panley01 yes, they have always admitted to using AI in detection. there is no good alternative to detecting this imagery that doesn't exists in the PhotoDNA database. but this incident didn't use AI.

@BlackbluueVT @dootskyre @panley01 Note the "machine learning" part of that article.

@CCShakeCat have you tried freeing up some space?

@BlackbluueVT I know, that's why I hate it. It's what got me signed out of Discord FIFTEEN TIMES. Nothing else does that I mean my phone, being android, tells me that I'm low on storage and that things may run slower, but it doesn't completely break down like Windows

Why is Winblows so self-destructive when you have low storage Apple- "Hey, you have low storage" Android- "Hey, you have low storage" Windows- "Goodbye cruel world"

@DrumstickPony Except AI didn't do any of those things in this instance.

This is what happens when you let AI be judge, jury, and executioner AI should never ever be allowed to moderate and this is why

I've seen a lot of people quote this post, falsely claiming that Discord's AI moderation is at fault. Please understand people, this is not related to their AI moderation. it's a file hash false positive. Don't be a boy who cried wolf; attribute the correct problem to the issue.

@CCShakeCat @OzwaldV001 which is fine. as long as you understand what you are being careful of

@BlackbluueVT @OzwaldV001 I'm still being careful

And my friends really thought it was complete BS Be very careful everybody

This weekend, our safety systems incorrectly triggered and banned around 200 accounts. Everyone affected has been reinstated. Here's what happened. A thread 🧵

@weldon1045 can to state the relevance of this image? where was a computer ever given direct management decision power?