-
Ever wanted an RSS feed of all your favorite gaming news sites? Go check out our new Gaming Headlines feed! Read more about it here.
-
We have made minor adjustments to how the search bar works on ResetEra. You can read about the changes here.
Activision will add real time in-game voice chat moderation in Call of Duty, AI powered detection - Live Today in MWII/WZ in NA, worldwide with MW3
- Thread starter Bishop89
- Start date
i guess that makes sense. It'd be impossible to have real mods in every game. And it says it flags things for actual review by real people
I swear to God I said <insert something that rhymes with some slur>!
Whenever video game accounts like Charlie intel use BREAKING esp in all caps for video games news, I just go like
Man, this isn't like a severe weather statement
It's call of duty news
Regardless this is good, hope it brings down hate speech
Man, this isn't like a severe weather statement
It's call of duty news
Regardless this is good, hope it brings down hate speech
Modulate
www.modulate.ai
Seems most are in beta (Spanish) or Alpha (everything non-English or Spanish)
Can you imagine only training AI off of COD voice chat?
*shudders*
If it's accurate enough I'm for it. It's just impossible to have humans do some tasks, moderating live chatter from millions of players is one of those impossible things.
Well it's training to detect speech and presumably "intent", not to generate anything
Now this would actually be incredible. People developing an entirely new language to get around moderation. There's already tons of gaming specific lingo and terms, might as well go all the way.
Lol, I hope they don't use that data for training because then we are all doomed
All it'll take is a couple thousand misfire bans and this whole thing will collapse like a house of cards.
Is this gonna end up like R6 Siege where people bait others into saying a slur?
Saw "Anyone remember the name of Eric Cartman's superhero persona?" all the time in chat.
Saw "Anyone remember the name of Eric Cartman's superhero persona?" all the time in chat.
Why do people always assume these systems take automatic actions when they flat out say it's meant to auto flag users
How many players play and use this feature? That sounds like it could be a lot of processing; whose paying for that?
I can understand banning text and report features etc but active voice moderation as a whole can't be cheap and it's Activision right?
Cool tech otherwise.
I think reports go to people to check for context.
Which makes me feel like they are also paying people to do this and listen in to conversations for context?
I can understand banning text and report features etc but active voice moderation as a whole can't be cheap and it's Activision right?
Cool tech otherwise.
I think reports go to people to check for context.
Which makes me feel like they are also paying people to do this and listen in to conversations for context?
If it's just flagging people and then a person is able to watch / listen that sounds like it could be good.
People are in party chat or have others muted to avoid the people these AI thingies are meant to censor.
I'm pretty sure this AI only understands english and that's going to be a problem.
People speaking other than english might get false banned due to this AI.
Something like this can happen:
https://gamerant.com/apex-legends-japanese-players-ban-run-word/
Imagine if two Norwegian dudes are playing CoD and talking to each other. Then they say something that AI thinks is racist, even though is not. That could easily lead to false bans.
There's also chance that people with speech impediment might get false banned.
However it seems that current system is just flagging at the moment, but knowing activision they might eventually get rid of human moderators and automate their jobs with AI.
I don't know why people in this thread are saying "great" when the risk of players getting false banned is high.
People speaking other than english might get false banned due to this AI.
Something like this can happen:
https://gamerant.com/apex-legends-japanese-players-ban-run-word/
Imagine if two Norwegian dudes are playing CoD and talking to each other. Then they say something that AI thinks is racist, even though is not. That could easily lead to false bans.
There's also chance that people with speech impediment might get false banned.
However it seems that current system is just flagging at the moment, but knowing activision they might eventually get rid of human moderators and automate their jobs with AI.
I don't know why people in this thread are saying "great" when the risk of players getting false banned is high.
"ā¦ā¦now it's a ghost town"
Will be so accurate now
Also I don't think I have spoken in a public lobby in a game in over a decade? Always in party chat instead, even when playing solo
Will be so accurate now
Also I don't think I have spoken in a public lobby in a game in over a decade? Always in party chat instead, even when playing solo
Let me guess it wont detect languages and spanish speakers will be banned for saying the word black
At first I was thinking of swearing and trash talk and was very confused. Like why bother in an M rated graphic military shooter. Meanwhile they sell mega bloks and still claim they don't want their games played by kids.
But hate speech makes sense.
If only it would work properly.
And I am wary of this type of thing in general anyway.
But hate speech makes sense.
If only it would work properly.
And I am wary of this type of thing in general anyway.
The complete meltdowns on twitter from people saying the game will die because they can't say the N word anymore make me happy
How exactly would you have human moderators listening in on every game?Imagine if it wasn't AI...
There's no need to speculate. The site for the service they use literally lists all the languages it supports/partially supports
Modulate
Seems most are in beta (Spanish) or Alpha (everything non-English or Spanish)
This is going to backfire in massive ways.
The AI will false ban people, and people are going to leave the game and go to something else.
The AI will false ban people, and people are going to leave the game and go to something else.
They definitely should use it for training just not for content generation. What better source would you have to detect toxic behavior