AI filtering

Question

This would use a blacklist and possibly an AI-enhanced swear detector.

If the filter detects bad stuff:

(This photoshopped image means that if something bad is detected, it can't be immediately posted, but clicking "Wait for Approval" will put it under review by a moderator.

Answer 1

Sorry for the late reply. It is very challenging for AI to detect hateful or bullying content that we don't allow. Personal information is also difficult to properly screen, especially when users try to hide it intentionally. Images are another problem.

Answer 2

boost

Answer 3

But then the filter would be filtering out wrong stuff that isn't even close to being inappropriate. Like on roblox users couldn't say simple words like 'hello' cuz it'd be blocked lol.

Comments

No, Roblox has a dumb filter because kids' accounts have to use a whitelist (meaning that any word not approved by Roblox will be hashtag'd).

My suggestion is that this uses a blacklist (blocks swears) with some sort of AI built into it.

Answer 4

Extremely cool, but chat would be like roblox chat?

“Hey, do you wanna ####?”

“WHAT”

“ITS NOT WHAT YOU THINK IT IS”

“WHAT DOES IT MEAN”

“####”

“I WONT ###### ####”

“WHAT DID YOU SAY”

Comments

It’d be more like PopJam filtering. (Which is awful, I’m glad PopJam shut down even tho KS had a PopJam account)

---

This is honestly very funny x3

Answer 5

Cool

Answer 6

Example anwer

Comments

How dare you

That is severely offensive

---

I love both this answer and your comment 

Answer 7

What is that on???? Thats cool!

Answer 8

That’s pretty cool. Really cool, actually.