Tune by Google

Tune by Google

AI-powered Chrome extension that hides toxic comments

4 followers

Tune is an experimental Chrome extension from Google's Jigsaw that lets people customize how much toxicity they want to see in comments. Set the “volume” of conversations on a number of popular platforms including YouTube, Facebook, Twitter, Reddit, and Disqus
Tune by Google gallery image
Tune by Google gallery image
Launch tags:
Chrome Extensions
Launch Team

What do you think? …

Miha Mlakar
Who determines what's a "toxic" comment?
Jason Sawtelle
OK @mihamlakar that's gonna be a strike for you!
Colin McDermott
@mihamlakar Certain keywords can be a bit of a giveaway. Not totally sold on this, but yeah I can imagine my life would better without having to read racist comments involving the N-word combined with multiple expletives for example.
Miha Mlakar
@colinmcdermott even if these platforms were littered with comments using the N-word (which I don't think they are) I believe it would be better for us to be able to see them and realize that as a society we have a problem. Don't you think so? I find the idea of a bot owned by a private company censoring the world for me really dangerous.
Colin McDermott
@mihamlakar - Okay here is a test for you. Go to LiveLeak.com and spend more than 1 minute reading the comments on the videos. There are 20 racist comments for everyone from a reasonable person. Tell me the tool wouldn't be perfect for places like this, for people like me that want to converse with normal people not racists. Dangerous to who? Are good quality sun glasses dangerous? How about waterproof trousers?
Anna Filou
On one hand we criticize filter bubbles, on the other we build things like this :)
Chad Whitaker
@anna_0x I tend to agree with you. But it does have a "volume" setting to at least filter out complete garbage (clear hate speech, trolling, etc.)? I'm experimenting with it on the lowest filter setting to see what it catches. Consequently, I now check and read all the filtered comments to see what it removed 😅
Anna Filou
@chadwhitaker the thing is, how does it determine which comments are actually complete garbage? How can we be sure it's not just really really rude truths being said, and not trolling? That'd be pretty hard to do.
Nikola Kozuljevic
@chadwhitaker @anna_0x or even how does it works with slang, quoting and paraphrasing? Language is highly dynamic (duh) and trolling with local slangs or idioms is hard to catch for humans even. If the algorithm (or whatever) relies on the human feedback to define what are unwanted comments (and it's likely it does) - that is a slippery slope. :) In any event - why not just include total comments black box. Isn't fake-instant gratification-bubble also harmful for young minds on the long run? :) (wondering why people even bother to read comments, barring on own content).
Aaron O'Leary
This is great does it have sentiment analysis built in or is it just a clean sweep of comments?
Arun Augustine
@aaronoleary yeah curious to know how this works.