© 2019 SourceMedia. All rights reserved.

AI's toughest challenge - Encouraging more women data scientists

In many respects, the artificial intelligence industry is responsible for driving new innovations with the potential to completely transform the way society operates. Through the work we see at Fountech, AI companies are generating innovative new solutions ranging from state-of-the-art cyber security systems to machine learning education programs.

However, we in the industry shouldn’t be complacent. While we are making great strides on the technological side of things, the industry is not immune to some of the widespread challenges facing the private and public sectors. This is particularly true of gender equality and diversity: two areas where there is still much work to be done.

A new report released by AI Now Institute in April 2019 shed some light on just how big an issue gender inequality – more than 80 percent of AI professors are men, while in leading tech companies like Facebook and Google, women comprise only a small minority of the AI research staff at 15 percent and 10 percent respectively.

With the AI industry growing at a rapid rate, it is vital to effectively address this issue of inequality now before the gender gap widens even further.

The unseen consequences of gender inequality in AI

We often assume that machines are neutral, but in reality, they aren’t. AI tools and technologies are themselves created by humans and feed on real-world data, which means they can inadvertently reinforce existing social biases. As AI plays a more important role in everyday decision-making, this will become ever more problematic should they accept current gender inequality as an established norm of society.

women in AI.png

This feedback loop was recently demonstrated by Amazon, when an experimental AI-driven hiring tool was used to identify suitable candidates. It was then discovered that the programme was biased in favour of men when it came to hiring technical talent, such as software developers.

By training ML algorithms to look for prospects by recognising terms that had appeared on resumes of past successful job applicants, the tool decided to downgrade resumes that contained mention of all-women’s colleges, or even contained the word “women’s’.

Importantly, the issue here is not the AI tool itself, but how humans have built the tool to learn and refine its selection process. As an industry, we have the opportunity to address these biases instead of embedding them in the next generation of technology.

What can be done about the gender imbalance?

It’s increasingly becoming clear why broader representation is needed within the AI industry, but how can organisations go about bringing more females into the industry?

Unfortunately, issues of bias in AI tend to most adversely affect people who are rarely in positions to develop the technology. That’s why more focus must be placed on encouraging women to seek positions in tech; as it stands, only 22 percent of AI professionals globally are female, compared to 78 percent of whom are male.

Driving change must begin through education – namely, encouraging women to pursue higher education within traditionally male-dominated fields like computer science and engineering. Across the UK, only 15.8 percent of undergraduates in STEM fields are women.

Through collaborative initiatives between public and private organisations, there is the opportunity to spread awareness and knowledge of the AI industry to a greater number of women deciding what to study, in turn breaking down the stereotype of tech disciplines as male dominated areas.

To ensure that AI is developed inclusively, workplaces must also place greater value on training and supporting their female workforce to stop the huge flow of women leaving the field.

Currently, it is estimated that almost half of the women who go into technology eventually leave the field – more than double the percentage of men who depart. We need to question why this is the case, and if necessary, look at ways of changing the current work culture of AI organisations. More generally, helping women pursue AI careers requires dedicated programmes and a pathway to long-term leadership roles.

For example, one of Fountech’s core objectives is addressing gender inequality. This is why as a business, we strive to hire and retain female talent, while at the same time offering advice for young graduates interested in a career in AI.

Make no mistake – I fundamentally believe AI has the potential to tackle pressing global issues, yet in order to reap the benefits, it is important we address issues of gender inequality now. Doing so will set a benchmark for other industries to follow.

For reprint and licensing requests for this article, click here.