Published 12:25 IST, September 22nd 2020
Twitter investigates as users point out alleged AI bias against women & people of colour
A Twitter user discovered that the image preview and cropping algorithm on the social media app was so biased that it picked white faces instead of black faces.
Advertisement
A Twitter user had alleged that he had discovered that im preview and cropping algorithm on social media app was biased and picked white faces instead of black faces. user tweeted several experimental pictures to prove his claims while sharing that professor Anima Anandkumar, who is director of AI research at nVidia, was first one to point out bias in 2019. He also suggested people follow Joy Buolamwini who is founder of Algorithmic Justice League.
Joy's organisation focuses on raising awareness about impact of Artificial Intelligence and believes that 'AI systems can amplify racism, sexism, ableism, and or forms of discrimination'. Twitter user named ar activist and book writer, Ruha Benjamin, who speaks about racism in techlogy through her book 'Race After Techlogy'.
Advertisement
While responding to user's experiment and discovery of racial bias, Twitter said, "We tested for bias before shipping model & didn't find evidence of racial or bias in our testing. But it’s clear that we’ve got more analysis to do. We'll continue to share what we learn, what actions we take, & will open-source it so ors can review and replicate."
Twitter's Chief Design Officer, Dantley Davis also responded to controversy in an attempt to explain situation.
Advertisement
User calls Twitter Biased
In order to understand racial bias in Twitter algorithms, a user posted a picture of Mitch McConnell and Barack Obama. In im preview, algorithm picked picture of McConnell instead of former US President Obama. user furr shared ar picture of two in order to understand if reason behind im preview could be 'colour of tie'. He first posted a picture of Mitch McConnell in a red tie and Barack Obama in blue which he later reversed using a picture of Obama in red tie and McConnell in blue but results remained same.
Advertisement
Twitter user Tony n shed light on racial bias by sharing a study published in Nature which examines use of software and algorithms which are widely used in US hospitals to allocate healthcare to people. study points out that use of this software leads to racial bias and discrimination against people of colour. In ar tweet, user pointed out that all people who attributed racial bias tp 'contrast' by editing pictures and raising contrast on Barack Obama and decreasing it on Mitch McConnell's pictures are completely missing out on point and core problem.
Advertisement
This experimental post came soon after a University manr from Vancouver, Colin Madland was helping out a faculty who faced a problem with video calling app Zoom which made head of man disappear during calls. Colin claimed that facial detection algorithm of app was so bad that it erases black faces. face of man was only visible when a pale globe was placed in background. Tweeting about it, Colin Madland said, "Turns out @zoom_us has a crappy face-detection algorithm that erases black faces...and determines that a nice pale globe in background must be a better face than what should be obvious."
Advertisement
Earlier in 2019, Professor Anima Anandkumar had pointed out that Twitter AI is -biased because when a picture of men and women is tweeted it simply puts focus on men instead of women and often crops out face of women in im preview. Following this, several users posted similar pictures and colls as well as screenshots to prove that bias is real till date. re were or Twitter users who called it AI misogyny while pointing our bias against women and people of colour.
12:25 IST, September 22nd 2020