Today, facial recognition processes on social media are driven by expert facial recognition artificial intelligence routines. Although artificial intelligence is generally thought to become smarter by learning over time, sometimes big setbacks can happen. The most recent example of this occurred in a Daily Mail message. Facebook artificial intelligence identified black men as “Primates” in a video post by the Daily Mail.
You can watch the video here, and on the right, you can see the erroneous message that is the subject of the news.
Facebook spokesperson Dani Lever apologized for this mistake, which she described as “unacceptable”, and said: “As we’ve said before, while we’ve made improvements to our AI, we know it’s not perfect and we still have a long way to go. We apologize to anyone who may have seen these disturbing suggestions.”
The studies that are the subject of improvements that Lever mentioned are a huge dataset that was made available by Facebook in April. This dataset has been opened for use in artificial intelligence research and education.
At the end of the day, it is possible to say with confidence: It is a fact that artificial intelligence is not smart enough yet…
36586
