TIMNIT GEBRU AND THE FIGHT FOR EQUALITY!

Timnit Gebru is an Ethiopian computer scientist and AI researcher born in 1983. She studied electrical engineering and did a Ph.D in computer science. She is known for her work on bias in artificial intelligence and co-founding the group Black in AI. In December 2020, she was fired from Google after raising concerns about the company's minority recruitment policy and approach to AI ethics. Her firing sparked a backlash from the AI community and renewed calls for diversity and inclusion in the tech industry.

A TIMELINE OF RACIAL INEQUALITY

1930's
REDLINING

Redlining was a discriminatory practice where services like mortgages were denied to residents of centrain areas based on racial or ethnic compositions. Although outlawed in 1968, long-term economic and social consequences like segregation and wealth gaps are still visible today.

redlining

1950's
SHIRLEY CARD

This card featured a white woman and was used to calibrate "correct skin tone" and colors in early photographs. Photography was not adapted for darker skin and led to poor representation of people of color in the domain of photography.

shirley-cards

1980's
MEDICAL PULSE OXIMETER

Pulse oximeters measure blood oxygen levels and have been lately found to be less accurate on people with darker skin tones. This leads sometimes to innacurate diagnosis and treament, making decisions that can sometimes be a matter of life or death! In the past, medical developpements weren't often tested on people of color or made by people of color, leading to privileged medical system for white people.

pulse-oximeters

20th century
AUTOMATED HIRING SYSTEMS

Automated hiring systems using algorithms to screen job applicants began developing in the late 20th century and are increasingly used in the 21st century. These algorithms can perpetuate biases present in the data they are trained on, for example if someone is from a foreign country or if their name is a non-white sounding name, these people will have less chances to be picked compared to a white person.

automated-hiring-systems

20th century
FACIAL RECOGNITION

Developed in the late 20th century and becoming prevalent in the 21st century, facial recognition technology is used today in many domains for example security, healthcare, education, entertainment and communication, and even travel. This technology has been found to have higher error rates for darker skin tones, often putting aside the importance for colored people to be represented in the technical and scientific system of our societies.

facial-recognition

2020
TIMNIT GEBRU FIRED FROM GOOGLE

After raising concerns about Google's minority recuitment policy and approach to AI ethics and their lack of tranparency, Timnit Gebru was fired from the Ethical AI team. Shortly after, her collegue was fired after searching to defend Timnit.

timnit-gebru
twitter
Timnit on Twitter

More on Timnit's work:

Black in AI
Gender Shades