Published , by Chris Jarrard
Published , by Chris Jarrard
Sometimes we may have good intentions, but end up failing to be of much help on account of perspective. Intel appears to be having a similar moment with their new Bleep audio filter tool. It uses the power of AI to remove potentially offensive language from voice communications.
Bleep works with other apps on your PC to act as a virtual moderator of sorts. Intel said it “uses AI to detect and redact audio based on user preferences.” An approach like this makes sense if you have aspirations of filtering all offensive content from your PC when your kids use it while still having the option of allowing adult language after the kids go to bed.
What Intel’s Bleep options screen actually offers is a bit puzzling. The UI is loaded with categories for the various types of potentially offensive language with each given a slider. You can tinker with how much filtering is applied to categories such as “Aggression,” “Misogyny,” or “White Nationalism.” I’m not fully clear on the type of use case that would benefit from only permitting 25% White Nationalism, 75% Ableism, but blocks all Aggression. Unlike all the other choices, the n-word gets only a full toggle.
The ability to design and construct my own personal hate speech filter as if it were a Chipotle burrito is notable from a technical standpoint, I guess. The usual amount of garbage that comes over public voice communications in video games is already so bad that many players disable voice capability within any new games they acquire. There is a need for technology like is seen with Bleep, though Intel may need a few more passes at the presentation side of things.