Not very Cres, but the schadenfreude is delicious.
It seems that Facebook’s AI committed what the company dubs an “unacceptable error”. The company’s AI followed up a video from a British tabloid featuring Black men with a prompt that asked users if they would like to “keep seeing videos about Primates”.
The video in question dated back to June 27, 2020. It came from The Daily Mail, and featured clips of black men getting into altercations with white civilians and police officers.
This blunder caused the company to disable the AI feature responsible, while apologizing and stating that they would not allow it to happen again.
As we have said, while we have made improvements to our A.I., we know it’s not perfect, and we have more progress to make. We apologize to anyone who may have seen these offensive recommendations.
Dani Lever, Facebook spokeswoman
The irony of the situation? The AI was technically correct (the best kind of correct), in the worst, most racially insensitive way possible. All humans are, in fact, primates, and we typically consider ourselves to be top of the heap when it comes to intelligence. The AI probably didn’t reason the suggestion out that way, but there is an irony to it all.
Needless to say, however, nobody else seems to have seen it that way, and even my first response to finding out Facebook’s AI managed to step in it (in what has been an ongoing parade of racial blunders) was pretty much this:
Source: New York Times
About Author
