Face recognition biased against Black women and girls

Wow! So much has happened since my birthday in September. Yeah, I know, I know. A terrorist attack happened on my birthday over a decade ago. But I was just thinking about the last 4 weeks since Sept. 11, 2018. Aretha Franklin joined the ancestors and a preaching patriarch tried to bumrush her homegoing service. That man remains in the White House with his Twitter fingered small hands. And Justices Sotomayor and Ginsberg have to deal with two asshats sexual abusers sitting on our Supreme Court. Don’t be discouraged. Those are facts. Not alternative facts. There is work to do when the going gets tough!

Any women who are marginalized are at a greater risk at losing jobs where employers use AI to hire, according to Microsoft Post Doctoral Researcher Timnit Gebru. We must begin to pay much more attention to the unintended consequences of technology. We all expose ourselves to consequences positive AND negative online. But those from marginalized groups, like trans- or cis-gendered women and girls of color, people with disabilities, and others are far more at risk than generally speaking White men, White women, or even White children of either sex.

Based on research from several studies, U.S. Senator Kamala Harris (D-CA), along with other senators, wrote to the FBI, FTC, and EEOC to express concern for mounting evidence that face recognition algorithms perpetuate biases of gender, race, and age.

This is of interest in my work as I begin to write a book about the harms of YouTube’s platform for young Black girls who produce musical content on its platform that also profits emerging and existing musicians and any affiliated music and tech companies.

The concern here is that these technologies may violate civil rights law and be “unfair and deceptive” to users, creators, and employers who benefit from using the data collected by tech companies as part of their hiring processes.

I constantly warn my students that increasingly employers expect college grads to willingly provide their Facebook password for inspection, which is a violation of their freedom of expression in my mind. Graduates do not have to agree to such requests but often do. What should happen is that they challenge prospective employers. If you make me an offer, I’d consider it but not before then.

So much of what people do in one role is not appropriate in another. This is called “role strain” by sociologists or “colliding and collapsing contexts” in social media researchers. For example, what you do in a room or online with your lover or sexual partner will often never be appropriate behavior with a parent or elder. When it is observed out of context, trouble comes in the form of misconceptions. Our online behaviors are being mis-read by the wrong viewers. This can easily happen with data collected from facial recognition software and AI.

The scholarship of a handful of women have brought so much detail to these areas of research that I almost want to assert that women dominate work aimed at understanding unintended consequences of tech. The work of danah boyd, Alice Marwick, Amy Bruckman, Nancy Baym, Safiya Nobile, and Simone Browne have been on my reading lists. I also love Christian Fuchs work as well.

Staying on top of new media research as a technomusicologist, borrowing the identification from fellow ethnomusicologist Wayne Marshall, takes time. But I hope my next book offers something to the disciplines of music. The primary book that I know of thus far is Kiri Miller’s books on YouTube. I’ve been a devotee to the work of Mike Wesch when it comes to YouTube, but haven’t truly carved out my own voice fully. The time has come.

Jill Walker Rettberg immersed herself in studying the app formerly called Musical.ly, which recently was bought by another company and is now called Tik Tok.

I’ve been experimenting with online platforms with undergraduate students in my anthropology courses including studies of selfies on Instagram

…and UGC on YouTube.

My current research and writing still focuses on Black tween and teen girls on YouTube but my primary concern is around the harm that stems from music and music-related UGC on YouTube. Who benefits from the user-generated content of black girls who twerk on YouTube? Who jump on the latest trends in music that tends to be sexually exploitative and often created an ecologically unfit space that primes young girls of any ethnicity to manufacture their consent to be sexual objects online. What consequences stem from Black girls’ online fandom?

In ways, this work is a continuation of my earlier book on kinetic orality and music between the sexes where Black girls figure centrally in the culture. Voice and embodiment are prime ways of thinking in my work. So I am still after ways to help readers see the power of body and voice for marginalized groups.

Social media changes Black girls’ bodies and representations in ways that are not always tangible or visible to them or their guardians despite the empowering ways they intend to perform their self online. This demands inquiry and study. That’s my work. To werk the body of research in ways that helps us all understand its affordances and costs across the life cycle.

Related posts:

Every year in January, the small Swiss town of Davos becomes the highlight of the entire world where more than 2500 leaders from business, government, international organisations, civil society…

A update was released on 7 oct 2018 with many bug fixes and layout changes such as –. “7th October 2018 Update” is published by Famenik.

On this day three years ago, March 11, 2020, the sky seemed to fall over our heads, and all it took was a tweet. In retrospect, it was just another Wednesday evening spent at home with my girlfriend…