Society diversity

Encouraging women in tech is key to protecting society from AI bias – TechTalks

By Xiaoman Hu

Encouraging women in AI has never been more urgent. A study of World Economic Forum noted a gender disparity of 78% male versus 22% female in AI and data science. This disparity is not just a challenge within the workforce. It reflects a very nuanced issue that goes beyond a single workplace and which, if left unaddressed, will have very negative implications for society.

We’ve seen a lot of work to encourage girls and women to engage with STEM and address digital skills gaps at an earlier age than in the past. Yet now there seems to be less effort to support women as they transition from higher education to a sustainable career in tech. It is a challenge for the industry. But the real problem is that as AI becomes pervasive in everyday life, without a tech workforce that accurately reflects the structure of society, AI-based decisions are constrained by societal biases. and cultural limitations of their designers. The impact of such consistency in AI decisions and biases has already been seen in examples such as automating credit card and mortgage applications, resume screening and other areas. .

The industry’s challenge is not due to a lack of skills. Research from the turn institute suggests that women lag behind men in industry-relevant skills such as computing, data preparation and mining, general purpose computing, databases, big data, l machine learning, statistics and mathematics. Yet much of this is not due to formal skills, but rather to women’s confidence in declaring these abilities during recruitment and in the workplace. In the world of technology where hard skills are needed, soft skills are sometimes overlooked, but moving forward requires more emphasis on leadership and mentorship to build trust and nurture a hand – more diverse workforce. It is said that we must fight stereotypes from an early age, but a gap remains. For example, in the tech sector, women generally have higher levels of formal education than their male counterparts, but academic citations are fewer, suggesting there is a lack of trust in sharing. academic knowledge. the Turing Institute finds that only 20% of UK data and AI researchers on Google Scholar are women. Of the 45 researchers with over 10,000 citations, only five were women.

When I say that women need to have mentors and role models, I write from first-hand experience. It wasn’t until I won a mathematical modeling competition in college that I considered a related career. It inspired me to write a blog about machine learning algorithms. The easy-to-understand method used helped the blog garner over 5 million views and ultimately led to a career in programming. When I became a programmer and found myself working as the only woman in a room of men usually 10-15 years older, I struggled to identify with myself and realized the need of a community of like-minded people.

In April 2020, I started managing the operations of MindSpore, an AI framework developed by Huawei, just when it became open source. MindSpore is Huawei’s alternative AI framework to Google’s TensorFlow and Facebook’s PyTorch with comparable capabilities but 20% fewer lines of code. Launched in September 2019, it is endorsed by leading universities including Peking University, University of Edinburgh and Imperial College. Today, MindSpore has over 1.3 million downloads and an interactive community indicated by over 19,000 issues, over 52,000 pull requests and over 16,000 stars (the equivalent of an “I likes” among the developers).

In 2021, downloads of open source components increased by 73% year-over-year. With the rapid growth in global adoption of open source technology, the diversity of open source communities is also increasing. The MindSpore Women in Tech community emphasizes seminar-style meetings that provide a safe space for women to discuss the challenges they face in the workplace. Mentoring is important. For example, in 2020, when the community was in its infancy, a student at one of our events explained that she was getting good grades but worried about a career in programming. She sought advice from more experienced programmers and technical managers. By the time she graduated, she no longer had to worry and was able to choose from several offers. Not only did she feel more confident, but she was able to give back to the community by sharing her experience with new students, those who were now in the position she held the previous year. It’s experiences like this that will keep women in tech. When they stay, technology also benefits.

But encouraging women isn’t just about creating diversity within the industry to allow for a better gender balance. The benefits extend beyond the sector and touch on societal benefits. With the digitization of many traditional sectors, the ubiquitous nature of AI demands that it not only be effective, but also inclusive. Only by expanding the talent pool can we avoid decisions based on biased data. Building communities that actively encourage participation and diversity of voices is an important step.

Bias in AI starts with the initial formulation of problems. The questions are naturally constrained by the experiences of designers and programmers. This in turn has an impact on the quality of the data and how it is processed. So what will be the societal impact if there is not more diversity?

  • The user experience (UX) for women won’t be as intuitive if there isn’t more input at the design stage
  • Economic discrimination if the allocation of women’s CVs to lower paid jobs and access to financial resources will have a long-term impact
  • Societal resources will be distributed unfairly, whether they affect education, health care or even security
  • Women will lose their decision-making capacities for day-to-day fundamental decisions

So, in conclusion, now that our lives are digitally driven, we need to ensure that women can enjoy the benefits of technology for generations to come rather than being negatively affected.

About the Author

Xiaomi Hu

Xiaoman Hu is the Operations Manager of MindSpore Community