Facebook AI catches 95% of hate speech, still wants mods back in office

Facebook’s software systems are getting better at detecting and blocking hate speech on both Facebook and Instagram platforms, the company boasted today—but the hardest work still needs to be done by people, and many warn that the world’s largest social media company is putting them in unsafe working conditions.

Facebook says in the latest Community standards Enforcement Report that around 95% of the hate speech on Facebook is taken by algorithms before anybody reports on it. Users reported the remaining 5% of the 22 million or so flagged posts in the last quarter.

Facebook has emphasized that while its internal AI has made progress with several categories of content implementation, COVID-19 continues to have an impact on its ability to measure content. It also emphasizes that in its news release and in the call with the press.

Full-time Facebook workers who work for the company itself are ordered to work from home or even indefinitely until July 2021.

Rosen underlined in a call to the reporters that employees on Facebook, such as those who manage essential features in data centers, are being provided with strict security precautions, and personal protection devices, such as hand sanitizers.

Leave a Reply

Your email address will not be published. Required fields are marked *