With the continuous development of the Internet, the channels for obtaining information have gradually shifted from newspapers and television to social platforms and search engines. However, the storytellers behind them are still dominated by white men and a few tech giants. These social platforms and search engines are not created and run by diverse teams. As a result, the content they produce is also likely to marginalize and stigmatize minorities intentionally or unintentionally, which in turn further perpetuates the inherent prejudice and discrimination, forming a vicious circle. Meanwhile, due to the monopoly of tech giants and the mechanism of algorithms, information cocoons and echo chambers will appear, making diversified content hard to see.
The diversity here refers to the?
Firstly, diversity can refer to anything that makes us unique including backgrounds, race, gender, disability, and sexual orientation. (“What do we mean by diversity and inclusion? “, 2022) Moreover, diversity can also refer to the diversity of the content shown on the internet. This kind of diversity deficiency can often result in political factors for instance China prohibits the spread of content about other sexual orientations except for heterosexuals. Besides, according to Paul X. McCarthy’s article Grandstand platforms, this diversity can also refer to the number of companies or platforms that occupy the technology industry. Nowadays, big tech companies such as Google, Meta, and Amazon have dominated our life which leads to a market oligopoly. All of these different definitions of diversity are within the scope of this article and are prevalent in today’s society and culture.
The pervasiveness of diversity deficiency
The root problem — a lack of diversity in the workforce
The cultural homogeneity of the workforce in the field of technology has been a problem in the history of the internet and can bring a series of problems (Hafner & Lyon, 1998). In the early days of the internet, white, middle-class men held the power to develop the internet’s infrastructure which prevented the internet from developing more fully (Castells, 2001). For example, machine-learning skin cancer detection systems have been shown to fail to detect cancers in people with dark skin tones (Buolamwini, 2017). Word association systems, such as those used by Google, automatically suggest searches that associate “housewives” with women, reinforcing harmful gender stereotypes (Bolukbasi, 2016).
However, the white male continues dominating the workplace nowadays (Chang, 2019). By 2021, women made up just 29% of tech companies’ workforce, while minorities made up just 22%. (“Representation of gender and ethnic minorities in technology organizations in 2021”, 2022) At meta, the number of female employees was only half the proportion of male employees in 2022 (“Meta: gender distribution of global employees 2022”, 2022). In leadership positions, 57.6% are held by whites, while only 28.6% are Asian, and 4.9% are black or African American (“U.S. Meta employees by ethnicity & department 2022”, 2022). Therefore, females and people of other races and sexual orientations are more likely to be excluded from positions that affect the technological system. While tech giants recognize that a diverse workforce — by gender, race, age, and other social factors — is good for business, and have made many new initiatives and promises, which turned out to be cosmetic. Without diversity in key roles such as leadership and executive roles, similar problems will continue to arise. For example, the cyber security field is an area with low representation, with only 24% representation of women and minorities. (Figure 1) While the field of network security is highly related to content moderation, which means that women or minority groups are less likely to participate in the work of content moderation, which will further lead to the lack of diversity in the content transmitted.
Derivative problems — lack of diversity in the content
First of all, scholars point out that each platform has its own set of normative logic and governance (Li, H., & Chen, X., 2021), which is likely to result in a lack of diversity in the content they transmitted. Most of the platforms emerging in the West were built by white men at the outset (as mentioned in the previous paragraph), so developers may bring their whiteness to the digital platform algorithms and normative logic consciously or unconsciously, resulting in limited content. This phenomenon is summed up by Matamoros (2017, as cited by Li, H., & Chen, X., 2021) as “platform racism” to describe how racism is amplified and replicated by the governance of platform offering business models and the user culture associated with these platforms.
The media industry which produces content all the time and affects people’s values is also found to be extremely lacking in diversity of content. According to Luo (2016), mass media can improve the visibility of individuals or groups in society and influence people’s values and perceptions of the world. On the contrary, it can also lead to the marginalization, stigmatization, and discrimination of minority groups due to the lack of representation or misrepresentation of minorities. In 2018, only 8.8 percent of broadcast television programs included LGBTQ people. (Noah, 2016) The media portrays heterosexuality as the “normal” of society, while other sexual orientation is severely underrepresented. Even if there is a representation of different sexual orientations in the media, it carries serious stereotypes. For example, mainstream media promote the idea that gay men are more attracted to advertisements selling expensive and possibly feminine products because of their assumed attitudes and lifestyles. (Kates, 1999) Even if there is diversified content generated, it may not be able to be seen as the recommendation mechanism of the algorithm is more likely to create an echo chamber. Consequently, people who are inherently discriminating will largely only see content that matches their values.
A video released by the BBC Reel showed that around 45000 roles credited to actors in the UK, only 0.5% were played by black British actors between 2006 and 2016. The result of not seeing the representation of a group will lead to a self-imposed limitation on achievement and allow for negative perceptions to go unchallenged. However, the video also reveals an easy way to achieve “diversity” which can be done by the use of making black characters more similar to white ones. Methods might be for the black characters to have fairer skin, white parents, or friends, make them get higher social status and then convey likeability. This method will result in further shaping and upholding racist perceptions.
Secondly, the lack of content diversity is partly due to government regulation. For example, in China, works with homosexual content are considered taboo. These works include any video, films, TV series, and novel articles on the Internet. Article 8 of the General Rules on Content Censorship of Online Audiovisual Programs (2017) approved by the China Internet Broadcasting Association clearly states that pornographic or vulgar content or plots should be cut and broadcast, including abnormal sexual relations, incest, homosexuality, and other sexual acts. Therefore, this lack of diversity is difficult to address, because it is even written into the association and clause.
The context — tech giants’ monopolies and government control
As said by Paul X. McCarthy, the lack of diversity also refers to the fact that our lives are or have been and are dominated only by a few tech giants. As a start-up enterprise, it is difficult to enter this monopolized industry and gain loyal users. This lack of diversity in software use is not only due to monopoly but also due to political factors. For example, China has banned Google, Facebook, Twitter, and other apps from operating in China since 2010. As a result, Chinese users are forced to use only Chinese search engines and social software.
A vicious circle
Due to chronic structural oppression, the lack of diversity saturates the offline workplace, especially in the tech sector, which has resulted in the inadequate development of the Internet infrastructure, leading to the emergence of platform racism. This lack of diversity in the workplace can also lead to bias and discrimination against women or minorities in the media industry. Stigmatizing information and content in turn continues to deepen the structural oppression inherent in society, creating a vicious cycle. While companies and media platforms are trying to improve diversity to varying degrees, such as hiring more women employees and improving the representation of actors of different races, dealing with this issue requires detail, care and caution. A focus on increasing the representation of women and minorities without concern about the way of presentation is likely to only entrench stereotypes.
Bolukbasi, T., Chang, K. W., Zou, J. Y., Saligrama, V., & Kalai, A. T. (2016). Man is to computer programmer as woman is to homemaker? debiasing word embeddings. Advances in neural information processing systems, 29.
Buolamwini, J. A. (2017). Gender shades: intersectional phenotypic and demographic evaluation of face datasets and gender classifiers (Doctoral dissertation, Massachusetts Institute of Technology).
Castells, Manuel (2001) ‘The Culture of the Internet’, in M. Castells, The Internet Galaxy: Reflections on the Internet, Business and Society. New York: Oxford University Press. pp. 36-64.
Chang, E. (2019). Brotopia: Breaking up the boys’ club of Silicon Valley. Penguin.
Hafner, Katie and Lyon, Matthew (1998), Where Wizards Stay Up Late: The Origins of the Internet. New York: Touchstone, pp. 43-81.
Li, H., & Chen, X. (2021). From “Oh, you’re Chinese . . . ” to “No bats, thx!”: Racialized Experiences of Australian-Based Chinese Queer Women in the Mobile Dating Context. Social Media + Society, 7(3), 205630512110353–. https://doi.org/10.1177/20563051211035352
Representation of gender and ethnic minorities in technology organizations in 2021. Www-statista-com.ezproxy.library.sydney.edu.au. (2022). Retrieved 14 October 2022, from https://www-statista-com.ezproxy.library.sydney.edu.au/statistics/1256194/representation-of-gender-and-ethnic-minorities-tech/.
U.S. Meta employees by ethnicity & department 2022. Statista. (2022). Retrieved 14 October 2022, from https://www.statista.com/statistics/311853/facebook-employee-ethnicity-and-department-us/.
What do we mean by diversity and inclusion? | Victorian Government. Vic.gov.au. (2022). Retrieved 14 October 2022, from https://www.vic.gov.au/dpc-diversity-and-inclusion-strategy-2019-2021/what-do-we-mean-diversity-and-inclusion.
Tsika, N. (2016). Glaad. In A. Goldberg (Ed.), The SAGE encyclopedia of lgbtq studies (Vol. 1, pp. 464-464). SAGE Publications, Inc., https://dx.doi.org/10.4135/9781483371283.n173
Steven M. Kates (1999) Making the Ad Perfectly Queer: Marketing “Normality” to the Gay Men’s Community?, Journal of Advertising, 28:1, 25-37, DOI: 10.1080/00913367.1999.10673574
LUO Muyuan. 2016. Paradoxical Visibility: An Analysis of Homosexuality Representation in The Nine Gay and A House of Zanker[J]. Fujian Journal of Agricultural Sciences, 36(2): 215-241.