Image licensed from Adobe Stock Images.
Every year we celebrate International Women’s Day with affirmations of diversity, equity, and inclusion (DEI) of women across organizations and society. But when we look at the trend it portrays a different picture, a picture of gloom and underrepresentation. Gender biases persist. Take, for instance, the STEM workforce — women are still significantly underrepresented. The data from LinkedIn on members’ job profiles clearly shows that women account for just 29.2 percent of all STEM workers though they make up almost half of the total employment across non-STEM occupations.
This alarming data also points to the number of women pursuing careers in STEM dropping significantly one year after graduation, indicating the presence of gender bias in these fields. When it comes to artificial intelligence (AI), the story is no different. Businesses and companies today have a common thread — a lack of women in the AI field. According to the World Economic Forum report, the pace of progress is disappointingly slow in AI, as only 30 percent of women are currently working in this field, despite a six-fold increase in talent availability overall between 2016 and 2022.
AI is peeping everywhere in our lives today from the daily appliances we use in our homes to our devices, vehicles, workplaces, healthcare, banking, customer service, education, recruitment, you name it and we have it. In this age of generative AI, where AI can generate content and images on its own, the potential for AI to change labor markets across the world is immense, and unprecedented.
The integration of AI technologies across industries has brought about a significant impact on the working lives of women. Despite several challenges, it has opened up new opportunities.
The World Economic Forum’s Future of Jobs 2023 report highlights that AI and machine learning specialists are among the fastest-growing jobs. Apart from AI specialists, according to LinkedIn’s 2024 Jobs on the Rise report, some of the most in-demand roles include AI consultant, AI leader, or vice president of AI. Machine learning engineer and machine learning researcher are also some top roles that are in demand with skills like Python, TensorFlow, and Scikit-Learn.
But coming back to the question of how many women do we have in these emerging roles or leading positions? As per the 2023 IBM Institute for Business Value Women in Leadership study, the representation of women in C-suite and board-level positions is a mere 12 percent. This concerning figure is further compounded by the decline in the pipeline of women leaders. Specifically, women hold only 14 percent of senior VP, 16 percent of VP or director, and 19 percent of senior manager positions, which is lower than the numbers recorded in 2019.
Image licensed from Adobe Stock Images.
Though we see progress in promoting diversity, inclusivity and equal opportunities in the workplace, the ongoing presence of gender bias is concerning. On why women are still vastly underrepresented in the AI workforce, Maja J Matarić, Chan Soon-Shiong Distinguished Professor of Computer Science, Neuroscience and Pediatrics at the University of Southern California, says:
“Under-representation of women is part of the larger issue of lack of diversity in tech in general. Some large companies have begun to work hard toward more inclusion, but progress is slow.”
“It would be good to understand the statistics about representation better; averages are not very useful. What are the AI jobs being surveyed and what does the distribution of % look like for those roles and various groups that tend to be under-represented? This nuance is important because methods to be used toward change are not the same across the board, in fact, there is no such thing as an “average strategy” that can be effective. So we need more data and those data need to be made public,” adds Prof. Matarić, also the Founding Director of USC Robotics and Autonomous Systems Center and Director of USC Robotics Research Lab.
A Deloitte survey found that several women face biases in the AI field and judgment from male colleagues and they also face the consequences of not conforming to the established male-dominated norms. The challenges start early. According to the same survey, 78 percent of women respondents did not have an opportunity to intern in AI or machine learning related roles while graduating. About 58 percent of the respondents said they left at least one job because of discrimination between men and women. Additionally, 73 respondents considered leaving the tech industry completely due to unequal pay and the inability to make headway in their careers.
Image licensed from Adobe Stock Images.
Prof. Matarić shares her perspective:
Here are some things that are being done and need to be done more.
a) Improving educational content for AI to make it more appealing and accessible to learners across the board: this includes simple but pervasive issues like pronouns in educational materials (yes, this is a real thing and it makes a difference) and larger key issues like diversity in educators (people who teach AI must not all be from a small subset of the population) and role models (successful examples must be diverse to make the AI jobs both appealing and relatable).
b) Providing not only mentoring for underrepresented groups, but also advocacy within organizations. We have known for a long time that mentors can make a larger difference, but lately we have learned from studies that what is also necessary is advocacy within an organization for change to happen. For example, a wonderful candidate from an under-presented group who has received great mentoring (still a rare case, so we need a lot more mentoring as well) can remain unappreciated and unhired (though possibly interviewed, but that’s not enough) if there is no advocacy within the organization toward truly giving all candidates a fair unbiased chance and consideration.
c) Reaching out to under-represented community forums, meeting places, and events, and taking part in person, rather than just sending ads and emails. Tweets/emails/ads are easy to do and noncommittal while real-world action makes a difference.
d) Making work about debiasing AI more visible; we all know there is bias in big datasets, and making efforts toward de-biasing visible and giving them recognition and credibility makes AI overall more appealing. Similarly, for AI for non-profit-driven ends: showing how AI can make a difference in addressing world-scale challenges like climate change, poverty, discrimination, etc. draws in much more diverse interest than just a typical set of obvious (but less compelling) uses for AI.
In a joint panel discussion by UNESCO, IDB, and OECD held on International Women’s Day, ’22, to discuss how AI affects women’s work, it was asserted that it is crucial to support the education of women and girls in STEM fields. Additionally, there is an urgent need to refrain from gender stereotypes in workplaces. Finally, there should be more studies on how AI systems are affecting work in general and women’s work in particular.
It’s a given that almost all companies have a common hurdle — lack of talent. Therefore, it is crucial to encourage women to acquire the necessary skills to keep up with the changing landscape and not be left behind. It should start early in school by educating girls about the potential opportunities in AI, as this field encompasses many different roles, such as data science, product management, AI ethics and user experience. It goes beyond just science or engineering roles. It is important to highlight women AI leaders to showcase the opportunities that await women in STEM.
Image licensed from Adobe Stock Images.
Organizations, on the other hand, still have the opportunity to address the gender gap in AI by prioritizing on the following:
While it can be challenging to address a gender-biased issue that has not been measured, things seem to be looking up as companies are recognizing the benefits of gender diversity in AI. Many organizations are now realizing the importance of retaining women in AI, and are coming up with initiatives like mentorship programs, women-only training, and flexible work plans to lure more women into the field. However, according to a Deloitte survey, women working in the AI industry don’t believe that initiatives such as diversity programs alone will be enough to make the field more equitable or attractive to women.
Establishing an inclusive work culture requires consistent efforts to identify and eradicate gender biases and discriminatory practices against women in workplaces. Historically underrepresented women also deserve equal opportunities and recognition.
But what do women want? Simply put: to be treated equally as men in workplaces, educational institutions and society. A small action that brings about a real change on the ground is much better than talking or advocating about bigger changes, which largely remain unexecuted.
© 2024 Mand Labs. All Rights Reserved.
