Using a very clever methodology, this year’s Digital Method Initiative Summer School participants show how generative AI models like OpenAI’s GTP-4o will “dress up” controversial topics when you push the model to work with controversial content, like war, protest, or porn.
Continue reading “Generative AI’s ability to ‘pink-wash’ Black and Queer protests”Racist Technology in Action: AI detection of emotion rates Black basketball players as ‘angrier’ than their White counterparts
In 2018, Lauren Rhue showed that two leading emotion detection software products had a racial bias against Black Men: Face++ thought they were more angry, and Microsoft AI thought they were more contemptuous.
Continue reading “Racist Technology in Action: AI detection of emotion rates Black basketball players as ‘angrier’ than their White counterparts”Are you 80% angry and 2% sad? Why ‘emotional AI’ is fraught with problems
AI that purports to read our feelings may enhance user experience but concerns over misuse and bias mean the field is fraught with potential dangers.
By Ned Carter Miles for The Guardian on June 23, 2024
How generative AI tools represent EU politicians: in a biased way
Algorithm Watch experimented with three major generative AI tools, generating 8,700 images of politicians. They found that all these tools make an active effort to lessen bias, but that the way they attempt to do this is problematic.
Continue reading “How generative AI tools represent EU politicians: in a biased way”Podcast: Art as a prophetic activity for the future of AI
Our own Hans de Zwart was a guest in the ‘Met Nerds om Tafel’ podcast. With Karen Palmer (creator of Consensus Gentium, a film about surveillance that watches you back), they discussed the role of art and storytelling in getting us ready for the future.
Continue reading “Podcast: Art as a prophetic activity for the future of AI”Image generators are trying to hide their biases – and they make them worse
In the run-up to the EU elections, AlgorithmWatch has investigated which election-related images can be generated by popular AI systems. Two of the largest providers don’t adhere to security measures they have announced themselves recently.
By Nicolas Kayser-Bril for AlgorithmWatch on May 29, 2024
I am not a typo
It’s time to correct autocorrect.
From I am not a typo
People with commonly autocorrected names call for tech firms to fix problem
‘I am not a typo’ campaign is calling for technology companies to make autocorrect less ‘western- and white-focused’.
By Robert Booth for The Guardian on May 22, 2024
Fouten herstellen we later wel: hoe de gemeente een dubieus algoritme losliet op Rotterdammers
Het was te mooi om waar te zijn: een algoritme om fraude in de bijstand op te sporen. Ondanks waarschuwingen bleef de gemeente Rotterdam er bijna vier jaar lang in geloven. Een handjevol ambtenaren, zich onvoldoende bewust van ethische risico’s, kon jarenlang ongestoord experimenteren met de data van kwetsbare mensen.
By Romy van Dijk and Saskia Klaassen for Vers Beton on October 23, 2023
LET OP, zegt de computer van Buitenlandse Zaken bij tienduizenden visumaanvragen. Is dat discriminatie?
Discriminerend algoritme: Volgens een onderzoek discrimineerde het algoritme dat Buitenlandse Zaken gebruikt om visumaanvragen te beoordelen. Uit onvrede met die conclusie vroeg het ministerie om een second opinion.
By Carola Houtekamer and Merijn Rengers for NRC on May 1, 2024
AI detection has no place in education
The ubiquitous availability of AI has made plagiarism detection software utterly useless, argues our Hans de Zwart in the Volkskrant.
Continue reading “AI detection has no place in education”The datasets to train AI models need more checks for harmful and illegal materials
This Atlantic conversation between Matteo Wong and Abeba Birhane touches on some critical issues surrounding the use of large datasets to train AI models.
Continue reading “The datasets to train AI models need more checks for harmful and illegal materials”White supremacy and Artificial General Intelligence
Many AI bros are feverishly trying to attain what they call “Artificial General Intelligence” or AGI. In a piece on Medium, David Golumbia outlines connections between this pursuit of AGI and white supremacist thinking around “race science”.
Continue reading “White supremacy and Artificial General Intelligence”Racist Technology in Action: Outsourced labour in Nigeria is shaping AI English
Generative AI uses particular English words way more than you would expect. Even though it is impossible to know for sure that a particular text was written by AI (see here), you can say something about that in aggregate.
Continue reading “Racist Technology in Action: Outsourced labour in Nigeria is shaping AI English”TechScape: How cheap, outsourced labour in Africa is shaping AI English
Workers in Africa have been exploited first by being paid a pittance to help make chatbots, then by having their own words become AI-ese. Plus, new AI gadgets are coming for your smartphones.
By Alex Hern for The Guardian on April 16, 2024
The Great White Robot God
It may seem improbable at first glance to think that there might be connections between the pursuit of artificial general intelligence (AGI) and white supremacy. Yet the more you examine the question the clearer and more disturbing the links get.
By David Golumbia for David Golumbia on Medium on January 21, 2019
So, Amazon’s ‘AI-powered’ cashier-free shops use a lot of … humans. Here’s why that shouldn’t surprise you
This is how these bosses get rich: by hiding underpaid, unrecognised human work behind the trappings of technology, says the writer and artist James Bridle.
By James Bridle for The Guardian on April 10, 2024
OpenAI’s GPT sorts resumes with a racial bias
Bloomberg did a clever experiment: they had OpenAI’s GPT rank resumes and found that it shows a gender and racial bias just on the basis of the name of the candidate.
Continue reading “OpenAI’s GPT sorts resumes with a racial bias”OpenAI GPT Sorts Resume Names With Racial Bias, Test Shows
Recruiters are eager to use generative AI, but a Bloomberg experiment found bias against job candidates based on their names alone.
By Davey Alba, Leon Yin, and Leonardo Nicoletti for Bloomberg on March 8, 2024
Google Used a Black, Deaf Worker to Tout Its Diversity. Now She’s Suing for Discrimination
Jalon Hall was featured on Google’s corporate social media accounts “for making #LifeAtGoogle more inclusive!” She says the company discriminated against her on the basis of her disability and race.
By Paresh Dave for WIRED on March 7, 2024
What Luddites can teach us about resisting an automated future
Opposing technology isn’t antithetical to progress.
By Tom Humberstone for MIT Technology Review on February 28, 2024
LLMs become more covertly racist with human intervention
Researchers found that certain prejudices also worsened as models grew larger.
By James O’Donnell for MIT Technology Review on March 11, 2024
Gemini image generation got it wrong. We’ll do better.
An explanation of how the issues with Gemini’s image generation of people happened, and what we’re doing to fix it.
By Prabhakar Raghavan for The Keyword on February 23, 2024
Google’s Gemini problem will be even worse outside the U.S.
It’s hard to keep a stereotyping machine out of trouble.
By Russell Brandom for Rest of World on February 29, 2024
Google does performative identity politics, nonpologises, pauses their efforts, and will invariably move on to its next shitty moneymaking move
In a shallow attempt to do representation for representation’s sake, Google has managed to draw the ire of the right-wing internet by generating historically inaccurate and overly inclusive portraits of historical figures.
Continue reading “Google does performative identity politics, nonpologises, pauses their efforts, and will invariably move on to its next shitty moneymaking move”Racist Technology in Action: ChatGPT detectors are biased against non-native English writers
Students are using ChatGPT for writing their essays. Antiplagiarism tools are trying to detect whether a text was written by AI. It turns out that these type of detectors consistently misclassify the text of non-native speakers as AI-generated.
Continue reading “Racist Technology in Action: ChatGPT detectors are biased against non-native English writers”‘Vergeet de controlestaat, we leven in een controlemaatschappij’
Volgens bijzonder hoogleraar digitale surveillance Marc Schuilenburg hebben wij geen geheimen meer. Bij alles wat we doen kijkt er wel iets of iemand mee die onze gangen registreert. We weten het, maar doen er gewoon aan mee. Zo diep zit digitale surveillance in de haarvaten van onze samenleving: ‘We herkennen het vaak niet eens meer.’
By Marc Schuilenburg and Sebastiaan Brommersma for Follow the Money on February 4, 2024
Machine Learning and the Reproduction of Inequality
Machine learning is the process behind increasingly pervasive and often proprietary tools like ChatGPT, facial recognition, and predictive policing programs. But these artificial intelligence programs are only as good as their training data. When the data smuggle in a host of racial, gender, and other inequalities, biased outputs become the norm.
By Catherine Yeh and Sharla Alegria for SAGE Journals on November 15, 2023
Timnit Gebru says harmful AI systems need to be stopped
The labour movement has a vital role to play and will grow in importance in 2024, says Timnit Gebru of the Distributed AI Research Institute.
By Timnit Gebru for The Economist on November 13, 2023
‘Unmasking AI’ and the Fight for Algorithmic Justice
A conversation with Dr. Joy Buolamwini.
By Joy Buolamwini and Nabiha Syed for The Markup on November 18, 2023
This is how AI image generators see the world
Artificial intelligence image tools have a tendency to spin up disturbing clichés: Asian women are hypersexual. Africans are primitive. Europeans are worldly. Leaders are men. Prisoners are Black.
By Kevin Schaul, Nitasha Tiku and Szu Yu Chen for Washington Post on November 20, 2023
Barbie and the dark side of generative artificial intelligence
As Barbie-mania grips the world, the peppy cultural icon deserves thanks for helping to illustrate a darker side of artificial intelligence.
By Paige Collings and Rory Mir for Salon on August 17, 2023
White faces generated by AI are more convincing than photos, finds survey
Photographs were seen as less realistic than computer images but there was no difference with pictures of people of colour.
By Nicola Davis for The Guardian on November 13, 2023
WhatsApp’s AI shows gun-wielding children when prompted with ‘Palestine’
By contrast, prompts for ‘Israeli’ do not generate images of people wielding guns, even in response to a prompt for ‘Israel army’.
By Johana Bhuiyan for The Guardian on November 3, 2023
AI is nog lang geen wondermiddel – zeker niet in het ziekenhuis
Tumoren ontdekken, nieuwe medicijnen ontwikkelen – beloftes genoeg over wat kunstmatige intelligentie kan betekenen voor de medische wereld. Maar voordat je zulk belangrijk werk kunt overlaten aan technologie, moet je precies snappen hoe die werkt. En zover zijn we nog lang niet.
By Maurits Martijn for De Correspondent on November 6, 2023
AI is bevooroordeeld. Wiens schuld is dat?
Ik ga in gesprek met Cynthia Liem. Zij is onderzoeker op het gebied van betrouwbare en verantwoorde kunstmatige intelligentie aan de TU Delft. Cynthia is bekend van haar analyse van de fraudedetectie-algoritmen die de Belastingdienst gebruikte in het toeslagenschandaal.
By Cynthia Liem and Ilyaz Nasrullah for BNR Nieuwsradio on October 20, 2023
Joy Buolamwini: “We’re giving AI companies a free pass”
The pioneering AI researcher and activist shares her personal journey in a new book, and explains her concerns about today’s AI systems.
By Joy Buolamwini and Melissa Heikkilä for MIT Technology Review on October 29, 2023
Some image generators produce more problematic stereotypes than others, but all fail at diversity
Automated image generators are often accused of spreading harmful stereotypes, but studies usually only look at MidJourney. Other tools make serious efforts to increase diversity in their output, but effective remedies remain elusive.
By Naiara Bellio and Nicolas Kayser-Bril for AlgorithmWatch on November 2, 2023
AI was asked to create images of Black African docs treating white kids. How’d it go?
Researchers were curious if artificial intelligence could fulfill the order. Or would built-in biases short-circuit the request? Let’s see what an image generator came up with.
By Carmen Drahl for National Public Radio on October 6, 2023
Instagram apologises for adding ‘terrorist’ to some Palestinian user profiles
Parent company Meta says bug caused ‘inappropriate’ auto-translations and was now fixed while employee says it pushed ‘a lot of people over the edge’.
By Josh Taylor for The Guardian on October 20, 2023