New Google's AI Chrome Extension To Clean Up Internet Comments

Tune

Just don't read the comments! That's is quite simple right? And it has become a moto for netizen these days. Because otherwise if you take your time going through them, you will see some evil things. Recently Google has released a new tool with the objective to give users the power to select what they want to see in the comments - the concept is quite nice but how will it wield?

Tune

So this new tool from Google is an extension for Chrome called Tune. It shares the machine learning model with the Perspective ( API from Google that flags toxic contents ). Tune also can indicate harmful content but for daily usage, not for publishers or developers. The extension will work on all leading platforms like Disqus, Reddit, Facebook, Youtube, and Twitter and it also has the setting to lets the users choose how much they want to left ou

From Jigsaw Google's manager of product - CJ Adams, we got to know that we will be able to adjust the toxic comment volume - which would be just like a volume dial. They have also developed a "zen mode" which would turn the volume all down to filter off the toxic comments; otherwise, we could see all the comments. In the middle ground, you could customize on which comments you want to see.

How will it work?

You install it just like any other extension on Chrome. Go on the Chrome App Store and do the rest like you normally do. After the installation, it will give you the option for filtering toxicity - but you also have many other options for customized filters ( but this is still labeled as " Very Experimental"). By the way, their definition of a toxic comment is that a comment kills the conversation. But there are also option to turn off other types like sexually explicit, attacks on identity, insults, threats, profanity.

There will also be checkboxes for you to choose to filter out these comments. Even though they are not perfect, they still can filter out those comments that will punch you in the face and make you very angry. Any attempt to clean the web from these toxic content is very welcomed by me.

How the extension looks like

Surely, to filter them out with this extension doesn't equal to delete them from the web, they are still there but hidden from your view. The filtering is unique to your preferences. It means that if you want to filter all the mention to lewd content or sexual acts in the comment, that applies only to your online experience. And as we have seen so far, there will also be filters for sex-positive and educational content as well.

So take in to notice the risk that, while you want to sweep out toxic and adult content in your comments, you might also leave out empowering comments on various subjects. The AI still lacks the needed understanding of the complexities and nuances of human languages; this was also proven through many failures recently of AI-based moderation efforts on social networks.

This is just the first step

From Jigsaw’s Adams point of view, he confirms that Tune isn't fully developed yet, they have said that on the blog spot before. Sometimes, the AI still misses some toxics comments or hide non-toxic ones. Therefore, their team is working very hard to improve the product based on the received users' feedbacks.  We must also notice that Tune is an all-in-one solution for toxicity on the web, nor a defense solution against direct online harassment.

They are just trying to help out by giving the power to the user to control what they will see and won't see. We all work for a clean and toxic free web, And Tune is just a glimmer at that future world. First, it is only an extension so that it will be useless on the app. Second like all other AI, it still has many things to learn before it could remove all the toxic content and embrace the good ones.

And with the default setting, with the description that it will remove comments that make you go way - that's vague. Will this apply on the content that you disagree with and serve to echo the online chambers?