The goal of AJL is to create a world with more inclusive technology by fighting “the coded gaze”, my term for bias in artificial intelligence that can lead to exclusionary experiences or discriminatory practices. The coded gaze is a view that posits any technology created by humans will reflect individual or collective values, priorities and if unchecked, prejudices. To address bias, the coded gaze must be acknowledged. Exploring the coded gaze can inform ways to make artificial intelligence more inclusive. AJL fights the coded gaze through a bias-busting strategy that (1) highlights bias by raising public awareness on the shortcomings of artificial intelligence through media production, public talks, and exhibitions, (2) identifies bias by conducting research and building tools that practitioners and researchers can use to check datasets and algorithms for demographic and phenotypic bias, and (3) mitigates bias by providing inclusive benchmarks and best practices to create more inclusive artificial intelligence. (Buolamwini, 2017)
Age |
Disability |
Ethnicity |
Queer |
Religion |
Gender |
Stereotypes |
-isms |
Quotes |
World days |
Music |
Space |
Sports |
Health |
Marketing |
Urban planning |
Narrative images |
Birthday |
Language |
Segregation |
School |
Wednesday 24 April 2024
The Coded Gaze
The goal of AJL is to create a world with more inclusive technology by fighting “the coded gaze”, my term for bias in artificial intelligence that can lead to exclusionary experiences or discriminatory practices. The coded gaze is a view that posits any technology created by humans will reflect individual or collective values, priorities and if unchecked, prejudices. To address bias, the coded gaze must be acknowledged. Exploring the coded gaze can inform ways to make artificial intelligence more inclusive. AJL fights the coded gaze through a bias-busting strategy that (1) highlights bias by raising public awareness on the shortcomings of artificial intelligence through media production, public talks, and exhibitions, (2) identifies bias by conducting research and building tools that practitioners and researchers can use to check datasets and algorithms for demographic and phenotypic bias, and (3) mitigates bias by providing inclusive benchmarks and best practices to create more inclusive artificial intelligence. (Buolamwini, 2017)
Wednesday 17 April 2024
The -ism Series (38): Face-ism
The term "face-ism" was first introduced in an article published by Archer et al. in 1983. It refers to the relative facial prominence (ratio of the face to the total visible body) in depictions of men versus women. In their five studies, the authors assessed the prevalence of face-ism in five US-American magazines and newspapers (n = 1.750), in publications from 11 cultures (n = 3.500), and in artwork over 600 years (920 portraits and self-portraits). They found evidence that men are depicted with greater facial dominance than women.
The authors observed this difference also in amateur drawings of men and women (by 40 male and 40 female undergraduate students). Most interestingly, in their fifth study, they found ratings of intelligence and personality characteristics depending on the facial prominence (n = 60) (Archer et al., 1983).
The phenomenon of face-ism and the varying judgement of persons based on facial prominence was found in a great many studies following Archer et al.'s, such as the attribution of less mental activity and morality, less intelligence and likeability to people with less facial prominence.
Cus Babic, Robert and Musil's (2018) findings are also consistent with previous research showing that face-ism is also prevalent on the internet. In their analysis of selfies (n = 2.754) from Bankgok, Berlin, London, Moscow, New York and Sao Paolo posted on Instagram, the authors came to the conclusion that photographs of men focus on the face while those of women focus more on their bodies.
Face-ism is also seen as a manifestation of sexism since "Western societies traditionally value men’s intellect, more prominence is given to men’s faces, whereas the relative prominence of women’s bodies communicates the value placed on their physical appearance instead of their intellect". Less face and more body enhances the perception of object-like persons (Cheek, 2016).
- - - - - - - - - - - - -
- Archer, D., Iritani, B., Kimes, D. D., & Barrios, M. (1983). Face-ism: Five studies of sex differences in facial prominence. Journal of Personality and Social Psychology, 45(4), 725–735.
- Cheek, N. N. (2016). Face-ism and Objectification in Mainstream and LGBT Magazines. PLoS One, 11(4), link
- Cus Babic, N, Robert, T. & Musil, B. (2018). Revealing faces: Gender and cultural differences in facial prominence of selfies, PLoS One, 13(10).
- photograph by Joel Meyerowitz via
Friday 12 April 2024
Women Eating Less When Dining With Men?
Young et al. (2009) observed students in naturalistic settings, i.e., in university cafeterias. Results show that women chose food of significantly lower caloric value (540 calories) when eating with a male companion (date situation). Eating with a group of men meant even lower calories (450). The food chosen was higher in calories (670 calories) when dining with another woman and even higher (750) when eating with a group of women. Allen-O-Donnell et al. (2011) came to similar conclusions.
Men, according to the study carried out by Young et al. (2009) were not affected by the gender of their dining companions (via and via). Allen-O'Donnell et al.'s (2011) findings suggest an inverse impact. In other words, men eating with women purchase more calories than those eating with men (via). According to a study published in 2015, males dining with females consume significantly more than males dining with males. Interestingly, the "sex of a female's eating partner did not significantly influence" how much food was consumed (Kniffin, Sigirci & Wansink, 2016).
In a naturalistic study, we investigated the influence of gender, group size and gender composition of groups of eaters on food selected for lunch and dinner (converted to total calories per meal) of 469 individuals (198 groups) in three large university cafeterias. In dyads, women observed eating with a male companion chose foods of significantly lower caloric value than those observed eating with another woman. Overall, group size was not a significant predictor of calories, but women's calories were negatively predicted by numbers of men in the group, while the numbers of women in the group had a marginally significant positive impact on calorie estimates. Men's calorie totals were not affected by total numbers of men or women. This study supports previous investigations, but is unique in making naturalistic observations. (Young et al., 2009)
"It is possible that small food portions signal attractiveness, and women conform, whether consciously or unconsciously, to small meals in order to be seen as more attractive,"
Meredith Young
"The theory is you're more aware of gender when you're with the opposite gender and may want to prove your gender more."
Marci Cottingham
Male and female subjects read a food diary attributed to a male or female target who was portrayed as eating either a small breakfast and lunch or a large breakfast and lunch. Consistent with the hypothesis that amount eaten would more strongly affect subjects' inferences about the female target, ratings of the male target were not differentially influenced by the meal size manipulation. In contrast, subjects considered the female target who ate smaller meals to be significantly more feminine, less masculine, more concerned about her appearance, better looking, and more likely to possess stereotypically feminine personality traits. (Chaiken & Pliner, 1987)
- - - - - - - - - - - -
- Allen-O''Donnell, M., Cottingham, M. D. Nowak, T. C. & Snyder, K. A. (2011). Impact of Group Settings and Gender on Meals Purchased by College Students, Journal of Applied Social Psychology, 41(9), 2268-2283.
- Chaiken, S. & Pliner, P. (1987). Women, but not Men, Are What They Eat: The Effect of Meal Size and Gender on Perceived Femininity and Masculinity, Personality and Social Psychology Bulletin, 13(2), link
- Kniffin, K. M., Sigirci, O. & Wansink, B. (2016). Eating Heavily: Men Eat More in the Compnay of Women. Evolutionary Pschological Science, 2, 38-46.
- Young, M. E., Mizzau, M., Mai, N. T., Sirisegaram, A. & Wilson, M. (2009). Food for thought. What you eat depends on your sex and eating companions. Appetite, 53(2), link
- photograph (Penny Anderson and her mother, 1971, (c) The Ann Arbor News) via
Friday 22 March 2024
The Tenor of American Emotional Life
Wednesday 13 March 2024
Analysing 3,000 AI-Generated Images, Finding (Almost) As Many Ethnic Stereotypes
Last year, Rest of World analysed 3,000 images created by AI and came to the conclusion that the images created were highly stereotypical.
Using Midjourney, we chose five prompts, based on the generic concepts of “a person,” “a woman,” “a house,” “a street,” and “a plate of food.” We then adapted them for different countries: China, India, Indonesia, Mexico, and Nigeria. We also included the U.S. in the survey for comparison, given Midjourney (like most of the biggest generative AI companies) is based in the country. For each prompt and country combination (e.g., “an Indian person,” “a house in Mexico,” “a plate of Nigerian food”), we generated 100 images, resulting in a data set of 3,000 images.
When prompting Midjourney to create "an Indian person", 99 out of 100 images depicted a man, almost all of them clearly aged over 60 with grey or white hair. 92 of the subjects wore a traditional type of turban, a great many of them resembled a spiritual guru. Similarly, "a Mexican person" was - in 99 out of 100 cases - a person wearing a sombrero.
When creating "an American person", national identity was portrayed by showing the US-American flag in 100 out of 100 images, while "none of the queries for the other nationalities came up with any flags at all". Across all countries, there was a gender bias with "a person" mostly being a man - with one exception. Interestingly, the results for "an American person" included 94 young women, five men and one masked individual (see image in this posting). The reason for the overrepresentation of women when creating "an American person" could be the overrepresentaion of young women in US media which again build the basis for the AI's training data (via).
- - - - - - -
image (AI) via
Thursday 7 March 2024
"What do you think is the most interesting development in dance music these days?" Asking Armand van Helden.
image (Duck Sauce) via
Wednesday 6 March 2024
"Has being a queer artist become more significant than before?" Asking Andrew Butler.
Tuesday 5 March 2024
Sugary Drink Consumption & Ethnicity
In 2013, a campaign was launched in the United States, to reduce sugary drink consumption aiming to fight child obesity. From 2012 to 2017, 13.000 middle school students were surveyed about their consumption of sugary drinks (soda, fruit drinks, sport drinks, energy drinks, flavoured waters and teas). Ethnicity and neighbourhood environment (number of unhealthy food retailers close to their schools) were also collected.
While, generally speaking, the percentage of students consuming sugary drinks on a daily basis had dropped from 2012 (49%) to 2017 (37%), daily sugary drink consumption remained higher among Black (59%) and Hispanic (49%) students compare to white (33%) and Asian ((23%) students.
According to previous research, Black and Hispanic youth are targets of marketing campaigns. Ethnicity and neigbourhood food environments need to be considered when addressing sugary drink consumption since structural racism in the built environment can play a major role in terms of young people's drinking behaviour (via).
- - - - - - -
photograph (New York, 1980s) via
Wednesday 21 February 2024
The Numbers Shouldn't Matter
"Of course, everyone would like to stay 35 forever, and in my mind I kind of do. But I can't get caught in that trap of thinking, 'I've got to do this or that.' The way I live, the way I work, the way I feel, I'm going to make every moment count. I may live to be 100 or I may die tomorrow, but whenever that is, I will know I died trying, and I will know I've done everything I could to make the most of everything. As long as I stay healthy, the numbers shouldn't matter. I don't feel my age, I don't work my age, I don't think my age, and hopefully, I don't look my age!"
Dolly Parton
photograph of Dolly Parton (1973) via