A group of leading voices got together to share some key actionable insights for women, men, and businesses, to help amplify women’s voices and get one step closer to achieving gender equality for women in data science and AI space.
What exactly are we dealing with?
Welcome to the world of Artificial Intelligence (AI), where the possibilities are endless, and the future is bright. But despite the rapid growth in this field, gender inequality remains a persistent issue for women in data science and AI space. As illustrated by the World Economic Forum, women make up only 22% of AI professionals globally; only 14% of AI paper authors are women, only 18% of authors at the leading AI conferences are women, and just 2% of venture capital was directed towards start-ups founded by women in 2019.
Bias and discrimination against women are unfortunately not limited to the real world. They are also reflected in AI algorithms, which have been shown to perpetuate and even amplify gender-based biases. For example, AI used in hiring processes may discriminate against women by excluding them from certain job opportunities. A striking illustration of this phenomenon is demonstrated by Amazon's recruiting algorithm, which learned from historical data that technical roles were predominantly occupied by men. As a result, the algorithm penalized resumes that included the word "women". Similarly, Google's job ads algorithm exhibited discrimination by showing high-paying job ads more often to men than women.
These examples demonstrate the serious consequences of biased AI systems and highlight the importance of addressing these issues to promote fairness and equality. Synthetic data can also help to mitigate bias in AI systems that utilize tabular data. Bias in tabular data can occur when the data used to train the system is not representative of the population it is intended to serve, leading to unfair or discriminatory outcomes. Fair synthetic data can be used to supplement or replace existing data to increase the diversity of the dataset and reduce bias. By doing so, we can work towards creating a more inclusive and just society both in the real world and in the world of AI.
Putting our heads together to come up with key actionable insights to empower women in data science and AI
Women are underrepresented, and their voices often go unheard. However, there is hope on the horizon! MOSTLY AI recently held a virtual panel on LinkedIn to discuss how we can achieve gender equality for women in data science and AI space and make their voices count.
The panel was hosted by MOSTLY AI’s Chief Trust Officer, Alexandra Ebert, and consisted of a diverse mix of AI and data science experts, who shared their real-life experiences, thoughts, and insights on the issues we face. From Sadie St. Lawrence, Founder and CEO of Women in Data; Matthew Ziebarth, Co-founder and CTO of Ada Growth; Lisa Palmer, Chief AI Strategist and AI doctoral candidate; to Pedro Pavón, Global Policy Director for Privacy and Fairness at Meta; and Noelle, Global AI Solutions Lead from Accenture, each panelist brought a unique perspective to the table.
The discussion kicked off by highlighting the challenges that women face in data science and AI space. Lisa Palmer, who has worked for several big tech companies, shared her experiences with discrimination, underrepresentation in leadership positions, and being excluded from important decision-making processes. Other panelists echoed these experiences, highlighting how women struggle to be taken seriously in male-dominated fields, face microaggressions, and lack support from their male colleagues. They also pointed out their concerns about the gender pay gap and paid family leave.
However, the panelists also focused on solutions for achieving gender equality in the AI and data science space. They noted that it is crucial to have more women in leadership positions and to provide them with equal opportunities for advancement. Creating a more inclusive workplace culture, where women's contributions are valued and recognized, was also stressed. Mentorship programs, networking opportunities, and training programs were cited as essential support structures for women, along with creating safe spaces for them to share their ideas without fear of retribution.
In summary, the panel emphasized the challenges that women face in the AI and data science space and discussed concrete solutions for achieving gender equality. By creating a more equitable and diverse AI and data science space, we can pave the way for a brighter future. Here are some essential steps that individuals and organizations can take to eliminate gender inequality in the workplace.
Key actions to drive gender equality for women in data science and AI
- Know your worth and embrace your uniqueness to help overcome any self-doubt you may have.
- Speak up when you notice unfair treatment, such as unequal pay or unfairness around promotions. Find allies to support you.
- Focus on specific actions you can take to tackle challenges in your immediate environment. Small changes can make a difference!
- Negotiate creatively, and don't be afraid to highlight your caregiving responsibilities or breadwinning role.
- Make sure potential employers understand who you are and what you bring to the table before joining their organization.
- Self-reflection is crucial to understanding how societal pressure on men negatively affects women. In many societies, men are often expected to conform to certain gender roles and expectations. These may include being strong, aggressive, and dominant while suppressing emotions and vulnerability. Media, popular culture, and social norms often reinforce these expectations. However, these expectations can lead to harmful behaviors and attitudes toward women in the workplace. Through self-awareness and -reflection, men can begin to challenge these harmful attitudes and behaviors and work towards becoming better allies and advocates for gender equality.
- Address toxic masculinity and be intellectually and emotionally honest with yourself.
- Use your position of power to help women by listening to them in a non-defensive way, answering their calls for help, advocating for them, and sharing and promoting their work.
- Take an intersectional approach and be mindful that different groups of women experience inequality differently, based on historic inequities.
- Provide feedback to other men to disrupt behavior that leads to oppression or inequality. Men checking each other is a powerful way to create change.
- Make roles accessible to and desirable for talented women.
- Create visible growth paths and highlight them to attract women into roles.
- Check job descriptions for language that may put women off from applying.
- Ensure gender equality is part of your organization's performance management and have a concrete plan for executing it.
- Highlight successful women as role models for younger women.
- Promote women to leadership positions so they can advocate for the next generation.
- Allow women to exercise their power and raise concerns without obstruction.
- Provide flexibility, especially for part-time work, as a glide ramp for women returning to work after taking time off for care.
- Prioritize in-house talent before bringing in new talent or making layoffs.
These actions are crucial for promoting diversity and equality in the workplace. Individuals and businesses should take them seriously to proactively create a more inclusive and innovative environment.