Alphabet-made Chrome extension is designed to tune out toxic comments

Image: Tune

Alphabet incubator Jigsaw has released an experimental Chrome extension that’s meant to make our online browsing experience a little bit nicer by freeing it of toxic comments. The extension, called Tune, lets you choose to see varying levels of polite or aggressive comments. “Zen mode” turns off comments entirely, while “volume levels” ranging from “quiet” to “blaring” let through different amounts of toxicity (like attacks, insults, profanity, etc).

The open-source extension uses machine learning to determine how likely a comment is to be perceived as toxic. It uses Perspective, an API created in 2017 by Jigsaw and Google’s Counter Abuse Technology team, and it’s used by news organizations including The New York Times and The Guardian to experiment with online moderation. Below is an example of how Perspective sorts comments by toxicity:

Image: Perspective

Jigsaw is careful to note that Tune is an experiment that’s still in progress, and it can be inaccurate at labeling which comments are toxic. The extension “isn’t meant to be a solution for direct targets of harassment (for whom seeing direct threats can be vital for their safety), nor is Tune a solution for all toxicity,” the extension’s description reads. Rather, it’s meant to show users how machine learning can be used to improve discussions online.

So far, the extension is available for YouTube, Reddit, Facebook, Twitter, and Disqus comments. Filtered-out comments show up blank with a dot, which can be opened to reveal the actual comment. Using it on my own Twitter feed set to the lowest “Quiet” setting, it incorrectly labeled this “devastating Beto attack ad” that’s actually a parody video as toxic. But can you really blame the AI? Machine learning still has a ways to go before it can detect layers of irony, and maybe this serves as an example that choosing to ignore things we don’t want to see can ultimately do more harm than good — or just make us miss a funny video.


My safe space, my safe space! Won’t somebody give me a safe space!

As entitled as you might feel to inject your toxicity into other people’s lives, there are people who would like to read the comments, who don’t specifically want to spend time reading yours.

Nothing to do with a safe space, and everything to do with some people simply not gaining anything of value by reading your comments.

Fair enough. I thought it was a little funny though.

When the phrase "over hyped nonsense" (in the gif) is flagged as toxic, I see this as increasing the dangers of echo-chambers.
Imagine this being used on a anti vaxx page, any corrections or explanations saying that links to autism are nonsense would disappear.

The phrase "overhyped nonsense" is typically used by trolls. I mean… what does that even mean? Overhyped based on what standards? Where’s the explanation about how it’s nonsense? It adds absolutely nothing to any imaginable discussion. I’m sure there’s more to it than just cherry picking sentences with particular words.

the phrase overhyped nonsense is overhyped nonsense

I agree that it’s not really constructive but the point was more one who decides on what is "adding to discussuon"

there’s a way to present your argument that something is nonsense without it being a toxic comment. I think that’s the point.

If you are standing in a blizzard and have the option to take shelter in a warm building, you are going to use the building as a safe space to avoid physical harm. If you are on the internet and being bombarded by negativity, some people might want a safe space to avoid mental harm. Mental health can be just as important as physical health.

You could always just leave the site. Its like if you had the ability to teleport and you were caught in a blizzard but instead of teleporting out of it you looked for a barn instead

By the then the damage is done. Some people have trouble shaking off a personal attack or a toxic comment even once they leave the site. There is also a reason most people don’t do that now. The internet can be a place for good social interaction and it provides unique opportunities to engage with people. If something like this extension can help filter out the toxic comments for those who want to avoid that kind of thing, then I see no harm in that.

That’s fair. Best to just avoid comment sections notorious for toxicity, like YouTube. Of course there can always be a rogue comment in other places.

I have no problem with a word filter, but a ‘toxic comment’ is a very vague and subjective description.

Agreed, and as the article discusses, it will never be perfect.

the snowflakes dream.
see cap marvel issue.
review bombed the movie because she "hurts" my feelings

I think it’s important to protect ourselves from harm in a lot of ways, but to protect ourselves from thought?

I imagine there’s a lot I don’t know, but I’ve always felt taking in different perspectives, even vitriolic ones allowed me to better empathize with the world and with viewpoints different from my own. Block those viewpoints and people out, and you’d be surprised at what you may miss or may not learn.

You may be surprised to find someone unexpected as president, for example.

There is a world of difference between actual constructive comments that might change your way of thinking and most of the garbage that gets spewed across comment sections. That stuff actually suppresses real nuanced debate and discussion. It’s not protecting yourself from thought, it’s tuning out the noise that prevents thought.

totally agree. mindless, vitriolic trolling seems to invade nearly every comment section, this site not excluded. ill definitely be interested in this.

would be great if it worked with Firefox.

It sounds great on paper but I continue to mistrust anything that purports to filter thoughts for me. It can easily be twisted for one point of views gain.

This. Slippery slope and all that. Would rather just have a simple word filter.

I think nuance can have a little emotion and vitriol. Not every conversation needs to be dry to convey a point.

In general, I’m just worried about who or what defines "noise". The examples in the GIF were fairly general besides the obvious "F you" comments. I’m just concerned that things like AI, ML, and even upvotes enable people to follow a hive mind mentality, highlighting whatever people find most agreeable. The best discourse always comes from challenge and skewed viewpoints, and even perspectives that come from different levels of intelligence that may not have arguments that are as well constructed as yours.

The ideal of shutting out the true chaotic evil of comments is certainly something to strive for, but I feel a filter just ignores the problem rather than discovering where feelings like that come from and finding a way to address it.

I think the issue is that some people don’t really want discourse — they are just trolls looking to see if they can upset people. There is nothing to gain from that.

they are just trolls looking to see if they can upset people

People get upset with anything they don’t agree with. Have you ever seen a debate online where one side doesn’t label the other "trolls"?
Watch Arkangel (Black Mirror) because that’s where we’re headed, as soon as Google finishes R&D on their contact lenses.

It’s not often that we get people who are actually only here to stir shit, and they’re usually pretty open about it. What we seem to notice is that people don’t always to operate as if there are other people on the other side of that comment they’re angrily bashing out a reply to, because looking at a screen for extended periods of time while not seeing anything resembling life, well…it has a way of obscuring the social aspect of internet interaction.

View All Comments
Back to top ↑