Just a few weeks after replacing all these people with software, MS Thar News accused racists of racism by Jade Tharwal, a member of the UK band Little Mix, after MSN posted a story about her opening up about racism experienced at her school, an image of Little Mix bandmate Le-An Pink peer.
Tharwal was unaware that the news was set up by an algorithm on the MSN site and thrown into lazy journalism. “This shit happens to that Lighannipink and I think all the time that it has become an ongoing joke,” Thrillwall wrote on Instagram. It’s lazy journalism. It is ignorant, it is cruel. It annoys me that you can’t distinguish two women of color in a team of four. In the article, I even have images that follow a picture of Leim here so that they can’t say we’re not the same mixed race? Well done.
As reported by the Guardian, images with articles have been selected by AI software. As we have learned time and time again (or not apparently), AI and machine learning technology have racist problems. From racist soap distributors who will not distribute soap in their own car that you are more likely to drive if you are black because they do not recognize skin tone, there are countless examples of machine learning technology that does not work as it should because it has not been tested enough (or at all) with non-white people in mind.
In 2009, computer scientists at Stanford and Princeton tried to train computers on how to identify an object. To do this, they collected a huge database of photographs of everything from asparagus to watering cans. Then they have a lot of photos of people – photos. The result was ImageNet, a huge (and most notable) object-recognition data set, complete with man-made inbuilt bias and propagated by artificial intelligence (AI).
The software used by MSN was unable to distinguish between Tharwall and Pinot, two mixed-color women of color, suggesting it may have had similar problems. The problem was corrected by a human, but too late. In a bizarre twist to the story, the Guardian reports that the rest of the human MSN staff has been warned that the AI software may try to publish the Guardian’s article about its AI software racist mix, and if they should try to delete the story, they should try to delete the AI. May try to “overroll” and republish the story on the website.
The rest of the humans have already deleted stories published by algorithms that criticize the release of the Little Mix mix, revealing that it was published in the first place. Perhaps news books still need people above all else and it is necessary to deal with traditional racism first before moving into the future.
