The internet is a place where many things can happen. People use it to inspire, motivate, organize, inform, and debate. But like all powerful outlets with minimally regulated environments, the internet has also become a place where disagreements can quickly turn into verbal sparring and even harassment. The removal of face-to-face contact and the anonymity provided by a screen and keyboard emblazons some users to unleash on one another, or rather, troll one another.
In case you’re unfamiliar with trolling:
Since the rapid takeoff of social platforms and the recognition of deeply problematic issues like cyberbullying, trolling has become as ingrained in the culture of online users as any other part of the internet. Some platforms, such as Yik Yak, seem to exist for the sole purpose of trolling, and online users can now expect the comments section of any shared post or news article to be flooded with deliberately offensive, provocative, and unproductive comments.
Trolling Hinders User Experience
Though the problem of trolling existed long before, it’s seemed to sky rocket in the past few years. A deeply controversial presidential election, growing concern surrounding international affairs, and an infectious fear of terrorist activity have all had a pernicious influence on online communities. The nature of trolling has only been made worse by the widespread sharing of actual fake news, as well as a more recently led witch hunt that promotes the defamation of prominent and reputable news source.
This, perhaps, is the inspiration behind the development of Google’s newest API, Perspective (API stands for Application Programming Interface-a set of instructions that tell developers how to integrate to an existing piece of software).
What is Perspective?
Perspective is an API made for publishers to incorporate into their websites. It cross-checks user comments against a human-generated database of comments that have been listed as offensive. As a user is leaving a comment, Perspective compares it to the hundreds of thousands of offensive comments in the database and then delivers a “toxicity” score on a scale of 0-100.
As it applies to the API, toxicity is defined as how likely a comment is to offend another user or cause someone to leave a conversation:
Working from this definition of toxic and the databases full of examples of such comments, Perspective will give feedback to commenters and website moderators in real time, like this:
Publishers will receive notification of comments that reach their toxicity threshold, and can then decide what to do with them.
Is this Google’s version of censorship?
The answer to this question might take some time to reveal itself. The API, only just launched at the end of February, does give publishers the power to determine their own toxicity threshold. Publishers can decide on a toxicity threshold of comments scored at 59 or 99, and the actions taken in response to such comments are also made at the publisher’s discretion.
Rather than being an application that censors conversation by way of placing a higher value on certain topics, the goal of Perspective is to improve conversations; part of improving conversations is making sure users aren’t driven away from the conversations in the first place, and thus the basis of Perspective was born.
Who’s using Perspective?
Since Perspective is so new, it has only been launched in partnership with a handful of highly reputable publishers, including the New York Times. Having won more Pulitzer prizes (119, to be exact) than any other news organization and being a leading producer of content across all cultural spheres (politics, art, opinion, international affairs, and more), the New York Times makes for a perfect test pilot for Perspective.
For the New York Times and platforms like it, the hope is that Perspective will facilitate diverse discussion, expand viewpoints, and provide a safe and respectful platform for their readers’ voices to be heard.