Coded Bias: A Film Review
The Netflix film, Coded Bias, outlines the ever-growing presence of artificial intelligence facial recognition technology and its widespread impacts. The film follows Joy Buolamwini, a researcher at MIT, who discovers that facial recognition is biased against both females and people of color, stemmed from her own personal experience using the technology. However, the film goes on to explore examples of AI technology in use all over the world.
One particularly compelling scene shows a gentleman being fined for covering his face while walking past a van armed with facial recognition cameras. He is stopped by the police, who demand to see his ID and then fine him. Another shocking scene depicts crowds of people in Hong Kong sporting masks, hats, glasses, and other facial coverings while protesting and vandalizing security cameras. Not to mention that the technology often misidentifies people, as shown by another scene where a 14-year-old black boy gets stopped by police due to a misidentification using AI. It shows that this is, without a doubt, a worldwide issue.
This film teaches viewers that artificial intelligence is a reflection of the society in which it functions. The algorithms of AI will eventually, if not immediately, reflect both the conscious and unconscious biases of humans. A shocking example the film explores is Microsoft’s “Tay” chatbot. Tay allowed humans to teach it through interaction: “The more you chat with Tay, said Microsoft, the smarter it gets” (Source). However, the concept went up in flames rapidly, as the software didn’t even last 24 hours. People quickly taught it to be racist and mysogynistic. It “learned” terrible things and took them as fact. Evidently, the technology was heavily flawed, and it was discontinued. Another example the film describes is with Amazon. Amazon began using AI to comb through job applications, in an effort to speed up the hiring process. Unfortunately, the technology went on to reject almost every single female applicant, ruling out any application with the word “women” on it.
The film leaves viewers with unanswered questions. How can facial recognition and surveillance bias be fixed? How can it be regulated? And what is the line between using AI and privacy rights? It’s clear we still have a long way to go to perfect this technology.
