Why we need to restore gender balance in AI
Nikolas Kairinos explores how AI is adopting gender bias from a male-dominated tech sector and looks at what can be done to prevent this.
Nikolas is the CEO and Founder of Fountech.ai, a company specialising in the development and delivery of intelligent AI solutions for businesses and organisations.
If we consider the challenges currently facing the artificial intelligence (AI) industry, one undeniable and immediate issue that comes to mind is gender inequality.
A new report released by AI Now Institute in April 2019 found that only 18% of authors at leading AI conferences are women, and more than 80% of AI professors are men. Moreover, women comprise only 15% of AI research staff at Facebook and 10% at Google – a concerning finding given how big and influential these companies are on the technological front.
Women are clearly underrepresented in the sector, and the statistics from this latest report pose a significant question to consider – is AI at risk of perpetuating gender inequality rather than solving it?
The ripple effect of gender imbalance in AI
As someone who has worked in AI for over three decades, I believe every organisation involved in the industry has an active role to play in bridging the gender gap and ensuring that women are involved in shaping the future of AI. The need for diversity is particularly important in AI when considering the long-term impact gender inequality could have in shaping the next generation of technology.
Relying on AI tools and technologies that have been created predominantly by men risks perpetuating gender bias and stereotypes. While AI is not itself biased, human input can inadvertently train it to be. AI programmes feed on human and real-world data to inform its decisions, meaning it is at risk of absorbing societal bias.
Indeed, AI solutions are increasingly informing decisions in critical areas of our society, including sectors such as recruitment. A case study from Amazon offers the dire effects that deeply ingrained bias can have on computer decision-making.
Amazon recently experimented with an AI-driven hiring tool to identify suitable candidates for technical positions within the company, software developers for example. Soon, however, ingrained bias towards men was discovered. By training the algorithms to look for prospects by recognising terms that had appeared on resumes of past successful job applicants, the system quickly began to downgrade resumes that mentioned an applicant attending an all-women’s college. More concerning, it also started omitting resumes containing the word ‘women’s’.
What can we do to fix this issue?
It would be naive to assume this problem could be resolved overnight. Real change must inevitably first start with education – namely, women should be encouraged to pursue higher education in technological fields. Traditionally seen as a male-dominated field, there are currently far too few women pursuing STEM subjects such as computer science and engineering. According to PwC, only 15.8% of undergraduates in STEM fields in the UK are women. Naturally, this means the industry is already on the back foot, with a limited pool of skilled female candidates.
Organisations also have an important role to play in bridging the gender gap and inspiring more women to join the AI workforce. For one, a greater emphasis must be placed on training and nurturing a female workforce, especially as it is estimated that almost half of the women who go into tech, eventually leave the field. A necessary step for AI organisations is to ask themselves just why this is the case, and what can be done to increase the retention rate of skilled women workers.
Creating dedicated training programmes to improve tech skills, as well as placing more women in leadership roles, are some of the viable solutions which I believe will help females fulfil their potential and help drive change across the industry.
Together, we can ensure that AI ushers in a new era of gender equality where everyone stands to influence and benefit from the latest innovations. I hold a very optimistic outlook for the future, and it is positive to see companies like Fountech recognising the challenge before us. AI could be at risk of exacerbating gender inequality, which is why it’s time to act now.