Technology Hopes to Save the Internet From Trolls

Rather than abandon online comments altogether, as one ABC affiliate did earlier of this year, exhausted editors are putting their faith in the final frontier: Technology.

Rather than abandon online comments altogether, as one ABC affiliate did earlier this year, exhausted editors are putting their faith in the final frontier: Technology. Technology, they hope, will save them from trolls, those ubiquitous commenters who derail conversations in hateful directions. Trolls are a growing problem: In fact no less than three well-regarded editors and writers invoked the term to describe the current state of Internet discourse this month. "Comment Sections Are Wastelands Ruled by Trolls," declared Wired's Mat Honan today. "Trolls are just getting more and more aggressive and uglier," explained the Huffington Post's Arianna Huffington on Wednesday as she announced a new policy banning anonymous commenters altogether. (Some find this move misguided.) And in a post for the site earlier this month, The Atlantic's own Bob Cohn described comment sections as "cesspools of vitriol, magnets for haters and trolls and spammers."

Up until recently, media organizations relied on teams of moderators to weed out the most hateful of trolls — like the kind who make rape threats. That, however, takes more resources — read: eyeballs — than many companies can afford. (As of October 2012, The Huffington Post employed 30 full-time moderators.) Editors and media technologists are hoping that new platforms and algorithms will further discourage or disregard trollery. These higher-level commenting systems loosely fall into the following three categories:

The Enhanced Comment Section: Many commenting systems, like the ones that The Atlantic Wire and The Atlantic use, rely on algorithms to reward the best comments, based on favorites. In theory, the technology pushes the best comments to the top of the post and the trolls to the bottom. but that only works if the most-liked comment is actually the best comment. Plus, if a post gets linked on another, site, trolls can manipulate comments sections, pushing their trollery to the top.

The Comment Blog: Wired's Honan points to Kinja — Gawker's new platform that allows commenters turn conversations into posts — and Branch, which moves conversations to a separate website. "Both of these systems treat discussions as independent acts instead of afterthoughts," writes Honan. In theory, since it it takes more effort to post on these platforms, the content will be of higher quality. So far, that hasn't turned out to be entirely true, as Slate's Will Oremus points out. Upon visiting Kinja for the first time he came across the following comment on an Elon Musk photo:

The Pseudo Hidden Comment: The Atlantic Wire's neighbor, Quartz, has in-post annotations, which allows readers to leaves notes in the margins. The new blogging platform Medium also has "notes" in its margins. Although both of these systems move comments into the actual post, rather than below, they end-up hiding commentary more than traditional comments sections do. With Quartz, users have to click on the annotation to see the text; a Medium writer has to green-light a comment to make it visible to the public. That, however, means that it's harder to see criticism, especially on Medium. (The site acknowledges this, telling its writers. "Don’t only make positive notes public. By showing the commentary of people who don’t necessarily agree with you, you build credibility." It's unclear how often this actually happens.)

So far, no one of these set-ups have managed to magically send trolls back to the dark corners of the Internet where they belong. And it's not clear that outlawing anonymity does any good either. See: Facebook, which has a real name policy, and a whole lot of terrible.