Attempt to tackle racial bias long overdue say practitioners, but it’s not just about the equipment.
By Aamna Mohdin for The Guardian on May 28, 2021
Automated systems from Apple and Google label characters with dark skins “Animals”.
By Nicolas Kayser-Bril for AlgorithmWatch on May 14, 2021
From Siri, to Alexa, to Google Now, voice-based virtual assistants have increasingly become ubiquitous in our daily lives. So, it is unsurprising that yet another AI technology – speech recognition systems – has been reported to be biased against black people.
Continue reading “Racist Technology in Action: Speech recognition systems by major tech companies are biased”
Color of Change petition calls Google’s block on advertisers searching for social justice content “unacceptable”.
By Leon Yin for The Markup on May 4, 2021
They said Google’s decision to block advertisers from seeing “Black Lives Matter” and other social justice YouTube videos was the last straw.
By Aaron Sankin for The Markup on April 20, 2021
The company is considering how its use of machine learning may reinforce existing biases.
By Anna Kramer for Protocol on April 14, 2021
In this piece for Markup, Leon Yin and Aaron Sankin expose how Google bans advertisers from targeting terms such as “Black lives matter”, “antifascist” or “Muslim fashion”. At the same time, keywords such as “White lives matter” or “Christian fashion” are not banned. When they raised this striking discrepancy with Google, its response was to fix the discrepancies between religions and races by blocking all such terms, as well as by blocking even more social justice related keywords such as “I can’t breathe” or “LGBTQ”. Blocking these terms for ad placement can reduce the revenue for YouTuber’s fighting for these causes. Yin and Sankin place this policy in stark contrast to Google’s support for the Black Lives Matter movement.
Continue reading “Google blocks advertisers from targeting Black Lives Matter”
For a Markup feature, Leon Yin and Aaron Sankin compiled a list of “social and racial justice terms” with help from Color of Change, Media Justice, Mijente and Muslim Advocates, then checked if YouTube would let them target those terms for ads.
By Cory Doctorow for Pluralistic on April 10, 2021
“Black power” and “Black Lives Matter” can’t be used to find videos for ads, but “White power” and “White lives matter” were just fine.
By Aaron Sankin and Leon Yin for The Markup on April 9, 2021
In 1965, IBM launched the most ambitious attempt ever to diversify a tech company. The industry still needs to learn the lessons of that failure.
By Charlton McIlwain for Logic on December 20, 2021
The article’s title speaks for itself, “Your iPhone’s Adult Content Filter Blocks Anything ‘Asian’”. Victoria Song has tested the claims made by The Independent: if you enable the “Limit Adult Websites” function in your iPhone’s Screen Time setting, then you are blocked from seeing any Google search results for “Asian”. Related searches such as “Asian recipes,” or “Southeast Asian,” are also blocked by the adult content filter. There is no clarity or transparency to how search terms are considered adult content or not, and whether the process is automated or done manually. Regardless of intention, the outcome and the lack of action by Google or Apple is unsurprising but disconcerting. It is far from a mistake, but rather, a feature of their commercial practices and their disregard to the social harms of their business model.
Continue reading “Filtering out the “Asians””
Since 2017, Mozilla – the makers of the Firefox browser – have written a yearly report on the health of the internet. This year’s report focuses on labor rights, transparency and racial justice. The piece about racial justice makes an interesting argument about how the sites we see on the first page of a search engine are a reflection of the general popularity of these sites or their ability to pay for a top result. This leads to a ‘mainstream’ bias.
Continue reading “The internet doesn’t have ‘universal’ users”
The answer to that question depends on your skin colour, apparently. An AlgorithmWatch reporter, Nicholas Kayser-Bril, conducted an experiment that went viral on Twitter, showing that Google Vision Cloud (a service which is based on a subset of AI known as “computer vision” that focuses on automated image labelling), labelled an image of a dark-skinned individual holding a thermometer with the word “gun”, whilst a lighter skinned individual was labelled holding an “electronic device”.
Continue reading “Racist technology in action: Gun, or electronic device?”
Technology has never been colorblind. It’s time to abolish notions of “universal” users of software.
From The Internet Health Report 2020 on January 1, 2021
Google has fired AI researcher and ethicist Timnit Gebru after she wrote an email criticising Google’s policies around diversity while she struggled with her leadership to get a critical paper on AI published. This angered thousands of her former colleagues and academics. They pointed at the unequal treatment that Gebru received as a black woman and they were worried about the integrity of Google’s research.
Continue reading “Google fires AI researcher Timnit Gebru”
The Markup has published an overview of the ways in which algorithms have been given decisional powers in 2020 and have taken a wrong turn.
Continue reading “A year of algorithms behaving badly”
A Google service that automatically labels images produced starkly different results depending on skin tone on a given image. The company fixed the issue, but the problem is likely much broader.
By Nicolas Kayser-Bril for AlgorithmWatch on April 7, 2020
The situation has made clear that the field needs to change. Here’s where to start, according to a current and a former Googler.
By Alex Hanna and Meredith Whittaker for WIRED on December 31, 2020
The company’s star ethics researcher highlighted the risks of large language models, which are key to Google’s business.
By Karen Hao for MIT Technology Review on December 4, 2020
Timnit Gebru’s firing could damage Google’s reputation and ethical AI research within tech companies, industry leaders told Protocol.
By Anna Kramer for Protocol on December 3, 2020
Google en YouTube geven informatie van Wikipedia weer. Dat gebeurt klakkeloos, schrijft Hans de Zwart. Als informatie niet klopt of als Wikipedia misbruikt wordt, grijpt Google niet in, of veel te laat.
By Hans de Zwart for NRC on August 8, 2018
Searching Google’s ad buying portal for “Black girls” returned hundreds of terms leading to “adult content”
By Aaron Sankin and Leon Yin for The Markup on July 23, 2020
Proudly powered by WordPress | Theme: Baskerville 2 by Anders Noren.