We tested Google's new harassment tool on hateful right-wing comments
In today’s harsh political climate, there's no shortage of abuse and trolls — which is why it's great that Google released a new tool to combat hateful speech on the internet.
It's called Perspective, and the New York Times and other publications will use the algorithm to comb through comments to search for harassment and insults.
SEE ALSO: 8 things you can do instead of watch Milo beg for attention
It rates messages on a scale from "very healthy" to "very toxic" so that they can be automatically removed or sent to human moderators.
We took the tool on a test drive with some of the far-right's most offensive comments.
Image: OLIVIER DOULIERY/GETTY, THE WASHINGTON POST/GETTY, DREW ANGERER/GETTY, ALEX WONG/GETTY, ALBERTO E. RODRIGUEZ/GETTY, DAVID S. HOLLOWAY/GETTY
The tool is certainly a work in progress. It doesn't pick up some subtle phrases that are abusive and hateful without using profanity.
For example, while thoroughly disgraced alt-right troll Milo Yiannopoulos' comment about women scores as highly toxic, his Black Lives Matter remark doesn't register as that abusive. Looks like Google could use some help with its algorithm.
Additional reporting by Marcus Gilmer.
