“I think the Internet is broken,” Twitter co-founder Evan Williams told The New York Times last May. “I thought once everybody could speak freely and exchange information and ideas, the world is automatically going to be a better place,” he added, “But I was wrong about that.” In his interview, Williams talked about how technology and social media has reshaped, and at times amplified, the way hate speech is spread around the world.
In January, the ad agency POSSIBLE launched the Twitter platform “We Counter Hate” in an effort to combat hate speech and other forms of online abuse. Their objective is to turn retweets of hate into $1 donations that benefit a non-profit organization fighting the cause.
The problem: Ray Page, executive creative director of POSSIBLE’s Seattle office, explained that more than ever before people are using Facebook to spread live violence – suicides, beatings and murder – while Twitter is filled with trolls and other types of abusers. “Fake news, whether created for ideology or profit, runs rampant,” he added, referencing Williams’ New York Times article.
The approach: We Counter Hate’s approach is three-fold: pledge support, identify hate speech through machine learning, and deter it.
To support the project, individuals can sign up for a monthly pledge or a one-time donation. To recognize hate speech, POSSIBLE leveraged enterprise-level AI platforms, Natural Language Processing and Image Recognition APIs to build the platform and teamed up with Spredfast to moderate incoming messages and identify hatefluencers (Twitter users who distribute discriminating content). Once identified, a human moderator steps in to deter their messages by permanently pinning this tweet: “This hate tweet is now being countered. Think twice before retweeting. For every retweet, a donation will be committed to a non-profit fighting for equality, inclusion and diversity. #WeCounterHate https://www.wecounterhate.com/counteredhate.”
The undeletable reply makes it clear that each retweet of the message will equal a $1 donation to the organization Life After Hate, which was created in 2011 by former members of the American violent far-right extremist movement to help reform people from violent extremist groups.
Page explained that with moderating hate speech, a human touch is essential: “Not all racism and bigotry is created equal; it can be very coded and nuanced. Since our replies vary, the AI can learn and become a more sophisticated tool over time.”
“We Counter Hate” was launched on Twitter as the platform is a de facto social media tool for hate groups to spread their ideology. In 2017, Page and his team observed over 16 million hateful tweets, compared to 15 million baseball tweets, and a little over 11 million tweets about the Super Bowl and Game of Thrones.
The feel good: We Counter Hate takes today’s negative online behavior and makes it positive. As a powerful force for good, Page explained that the idea behind the project was two-fold: slow the spread of hate by making people think twice about what they’re retweeting, and leveraging current technology to reshape existing behavior.
“The strength of this initiative lies in the moment of pause when people see it’s an undeletable pinned tweet,” Page said. Some of their progress shows a 64 percent reduction in collective retweets compared to the author’s previous tweet history, with many users deleting the deplorable messages altogether. “Creatives, strategists and developers alike have a creative responsibility to change the perception for brands and people. It can be small or large but we should all use our gifted skill set to spread a message of positive change and make the world a better place,” he added.