It’s a divisive question. When KJR released this as the title for our panel event recently, we came up against strong opinions. Why is it just women? Why should men trust AI? Why should anyone trust AI? What do we even mean by ‘trust’. We were accused of taking too narrow of a focus, but in doing so, the questions erupted. The wheels turned in our attendees and peers alike, and the conversation has commenced. Just as we had hoped…

Our panel was held in conjunction with the Something Digital festival in Brisbane. We welcomed an intimate room, yes, predominantly female (on this occasion) to absorb the learnings and insights from three perspectives: the employer, Kelvin Ross from IntelliHQ; the mentor, Karolyn Gainfort from KJR, and the AI trailblazers, Beck Simpson from Maxwell Plus and Serena Mou, PhD student at Australian Centre for Robotic Vision. Here are some of the themes that arose:

The truth about biased data

Becks: Look at your data sets – are the outcomes skewed? If you put biased data in, you will get biased data out. If you know you’re working on something with previous bias, you need a process, so you don’t end up on the front page!

Karolyn: Do our teams have diversity? What risks do we face if they don’t? Organisations need to ask the question early. People generally have good intentions, they don’t set out to cause biased algorithms.

AI myths

Karolyn: It’s not just the math genius who can make an impact on the future of AI. Psychologists are going to be important, so are anthropologists. We need those research skills. Ask yourself, what is the contribution I can make to AI?

Women in AI

Serena: Don’t get us wrong, we need men in AI, but we need a balance. It’s time to shift the view of boy industries or girl industries.

Kelvin: In your team you want people who do think differently from you. In creating this program (Young Women Leaders in AI), I ask myself, do I even want to tell other employees about these amazing women! Developing and employing young women in tech is a real competitive advantage.

So, should women trust AI?

It’s a valid question. As are all the others that flew out of its shadow. You simply need to look at reports of biased data already implemented in real-world situations to appreciate this. The USA has had issues in algorithms relating to criminal intent. From COMPAS, accused of imposing a racial bias on convicted criminals likely to re-offend, to PredPol, designed to reduce human bias in policing but in actual fact may exacerbating racial bias, AI algorithms that don’t address bias could have very real, very dangerous implications. As for women? The best examples to arise have been in an employment setting, putting society back 50 years with an inherent bias towards the hiring of women. Cue the recently publicised Amazon case where ‘masculine language’, as favoured by the engineers, put women towards the bottom of pile.

While our panelists may not have definitively answered this (can anyone?), the event took important steps towards building an understanding of the risks and opportunities posed by AI in our society amongst a curious and pro-active crowd. Some of whom might just be the future leaders AI needs.

As a founding Platinum sponsor for the Young Women Leaders in AI program, the topic of AI bias and ethics is high on KJR’s radar. We’re proud to facilitate and contribute to these conversations. And while we might not have all the answers, yet, we hope that as a result, we can guide and support both young women and industry to think further about the future of AI and what that means for them.

The Young Women in AI program is backed by Australian Government funding from the Women in STEM and Entrepreneurship (WISE) program and applications for its inaugural intake of 75 young women from around the country are now open here. The program is also accepting expressions of interest from mentors and seeking corporate sponsorship. Please visit the website for more details.