 |
| Facebook reports a drop in hate speech |
Facebook responds to whistleblower Frances Hogan's turning the story into hate speech.
Jay Rosen, Vice President of Corporate Integrity, posted an article defending anti-hate measures on social media, and said the decline in hate speech isn't just due to the existence of such content.
Rosen said the prevalence of hate on Facebook has fallen about 50% over the past three quarters to 0.05% of content viewed, or about 5 in 10,000 views.
"The technology we are using to combat hate speech is not enough and we are wrong to deliberately distort our progress," he added.
"We don't want to see hate on our platform, nor from users or advertisers," Rosen wrote. We remain transparent about their removal. These documents clearly show that our work of integrity is a journey that spans many years. Although we are never perfect, our team has worked hard to develop our system, find problems, and create solutions.
The CEO claimed it was wrong to focus on removing content as the only measure. Rosen said there are other ways to combat hate and companies should be confident before removing content.
This means that you need to be careful not to accidentally delete content and restrict access to people, groups and pages that may violate the policy.
The company sometimes encounters problems because content is mistakenly flagged as hate speech and deleting the system can lead to more accidents. Likewise, when few people view a particular post, hate has limited effects.
Hugin noted in his testimony that the company had very few objectionable materials. If so, it will still be an issue even if only a few users view the material.
Facebook hopes to change history after testifying against them
The post appears to be in response to an article in the Wall Street Journal that said Facebook employees charged with sponsoring objectionable content on the platform do not believe the company can reliably filter it.
The Wall Street Journal report noted that internal documents show that two years ago, the company shortened the time it allowed human listeners to focus on hate speech complaints. He also made other changes to reduce the number of complaints.
This, in turn, helped create the impression that the company's AI was dictating the rules better than it actually was.
Rosen's response also did not respond to Hugin's allegations that the company refused to implement safer algorithms to reduce hate interactions.
The company can make progress in reducing hate. But that's not Hugin's view, he said, and the social media company hasn't done enough.