Candid app aims to increase online civility

Candid, a new app, uses a language processing system to analyze online content and help remove hate speech, threats, and slander.

First Look

The ability to be anonymous online has turned the web into a hostile environment for many users, but the creators of a new app are aiming to counter inappropriate behavior online. Candid, an anonymous chat app, uses a language processing system to flag inappropriate content online for removal, hoping to make the internet a somewhat friendlier place. 

Using an individual's "education, employment, interests and neighborhood," the app provides users with a personal feed of content, as Engadget reports. The app has rules that ensure privacy for its users, encrypting personal information in a one-way hash, an encryption tool, before they are received by the company, and giving a new, randomized username to each post.

In a bid to keep users' interactions positive, Candid gives users "badges" based on their online behavior, including "explorer" or "giver." Some Candid users, however, could risk being labeled "gossip" or "hater," depending on their activity. 

A "lot of factors contribute to getting the hater badge," a company spokesperson told Engadget's Andrew Tarantola, "including the number of negative comments and posts based on sentiment analysis, number of down votes a user gets and the number of posts from the user that were taken down. Posts that are taken down by the system are reviewed by a human."

The app also has a system in place to attempt to verify news content and take down inaccurate rumors. In another sign of the app's attempts to be a positive force, it will send a notification with information about local crisis helplines to users threatening self-harm. 

However, Mr. Tarantola concludes that he thinks the app takes unnecessary anonymizing steps, although he wrote that it is well-intentioned. 

Hostility on the internet has long been an issue, and gained even more attention over the past couple of days in the wake of "Ghostbusters" co-star Leslie Jones' receiving a large number of racist and sexist comments from Twitter users.

Ms. Jones began re-tweeting her harassers to call attention to the abuse. "I leave Twitter tonight with tears and a very sad heart," she tweeted. "All this cause I did a movie." The next day, Twitter banned Milo Yiannopoulos, a controversial free speech advocate, for his role in encouraging the harassment. 

Previously, artificial intelligence has reflected the worst of negative online citizenship when Microsoft launched an AI program called Tay, designed to chat like a teenage girl. Twitter users abused Tay's commenting ability to make her respond in racist, sexist, or otherwise offensive ways to their tweets, and Microsoft pulled "her" Twitter account after just a few hours. 

Yet simple solutions may make an impact on webizens' behavior. Reword, a Google Chrome extension, reminds user to reconsider using hurtful language on social media by putting a single red line through an offending word. Users can override it, but it's enough to make many reconsider, its creators say.