Series Six, Episode Seven: Tales from a female AI entrepreneur, capital raising, data bias, the ‘heroine’s journey’ and practical ways men can support

Julia: Hello, my name is Julia Streets and welcome to DiverCity Podcast, talking about equality, inclusion and diversity in financial services. 

As many listeners to the show will know, I travel the world hosting conferences, speaking at many different events about diversity and inclusion. Thinking about the world of financial services and also technology, on my travels I get to meet some amazing people. It is a great pleasure today to be sitting in the offices of Flamingo AI talking to Dr Catriona Wallace. 

Let me tell you a little bit about Catriona. Based between the US and Australia, Dr Catriona Wallace is the Founder and Executive Director of the artificial intelligence FinTech and ASX listed company, Flamingo AI. Flamingo AI is the second only woman led business ever to list on the Australian stock exchange. Catriona has been recognised by the Australian Financial Review as the most influential woman in business and entrepreneurship and achieved Advanced Australia’s highest award in technology and innovation. She has a PhD in organisational behaviour, technology substituting for human leaders, and she was recently inducted into the Royal Institution of Australia, acknowledging her as one of Australia’s most preeminent scientists.

Catriona is one of the world’s most cited experts and speakers on artificial intelligence, customer experience, ethics and human rights in technology and is also a philanthropist, human rights activist and a mother of five. Catriona, thank you so much for being on the show.

Catriona: Julia it’s such a pleasure.

Julia: I have so much to talk to you about because we had the great joy of being on stage, doing a fireside chat at the Women in Payments conference in Sydney. There were many things that we were beginning to talk about and, in our preparation, I was really fascinated with your conversation around the importance of diversity in artificial intelligence and technology. I’d love to hear your thoughts about why diversity and inclusion really matters.

Catriona: A good place to start is just to identify what we’re talking about when we talk about artificial intelligence. In its most simplest term, artificial intelligence is software that mimics human intelligence. If it’s able to learn on its own accord each time it does a task, we call that machine learning. This is now the fastest growing tech sector in the world. So $38 billion invested in it and it’s likely to be replacing 40% of administrative and service entry level jobs within the next five years. It’s the time now that we call the fourth industrial revolution where things are fundamentally going to change. AI has been likened to the same as when electricity came to the industrial era or in fact fire was invented. This is profound and when we’re thinking about how it’s being made and who’s making it, we start to see that there may be a few cracks in this otherwise quite utopian dream of the robots coming and helping us humans do more things.

Some of the challenges we’re starting to see is in fact that, when we look at his coding, the algorithms that will be running our lives, we see that there’s less than 10% of machine learning or artificial intelligence high tech engineers who are female. It is 90% male coders and predominantly younger male coders. We might start to think there’s a bit of a challenge in diversity in teams. However, that’s not really the biggest problem. What the problem is, is actually the diversity in the data and the way that the algorithms are being trained to behave. What we do know already is that the data sets that are historical datasets that are used to train AI and machine learning tools, are represented in what the past of humankind was. It’s actually not that favourable. It’s not a great thing that we just take data that represents some of the biases that we already have.

For example, in the largest data set in the world that’s used to train most big machine learning algorithms, is called ImageNet. It is plagued with all types of language and categorisation. Some examples of this is when it does image recognition, it pulls up photos of middle aged men who are single and classifies them as widowers. It pulls up photos of young women and classifies them as man traps. It’s got this already inherent bias and these massive data sets that have been coded by either well intentioned or otherwise humans that may or may not be aware of their own bias. The bottom line is the datasets that we’re using to train the machines that will run our world are full of bias. And that’s a real problem.

Julia: What can we do about it? Is it about awareness in terms of the construction of these data sets and do you see a shift in awareness that this needs to be addressed?

Catriona: Yes, the awareness of what artificial intelligence machine learning is in the general population, or even with business leaders, is still very low. In fact, my experience in working across large companies across the world is that the higher up you go in these organisations, the less the leaders actually know about artificial intelligence. Certainly at the board and executive level, there’s very limited knowledge of what AI is. Awareness is a big factor. It’s also a real challenge because the people who are coding the machines are engineers, they’re not statisticians. Even if they’re a data scientist, they might be a mathematician who can code, but they’re not statisticians.

Those of us who are statisticians actually know how to clean datasets and run statistical tests and make sure the data sets are cleaned and absence of any skew data or bias before we use them to train the algorithms of the machines. But this craft has been lost in the modern world. When there’s pressure on the vendors, pressure on organisations to be building AI and be building it quickly, what is the incentive for us to make sure that the data is without bias? There’s very little, and there’s also a complete absence of laws and regulation around this topic.

Julia: I know that one of the things you’ve been thinking about is where does the governance come in, not only in house within the organisations with those coders and hopefully statisticians tackling the challenge, but also the outside world. We live in a time where everybody wants to collaborate, we all want open APIs and we all want to be able to have our data flowing seamlessly from outside our organisations into the next. I know you’ve been thinking about governance and collaboration.

Catriona: I’m working very closely here in Australia with the Minister For Technology and also the Human Rights Commissioner. We’ve just put together, and the government has announced, a human rights and ethics framework for artificial intelligence as well as an AI roadmap for Australia. Now that’s great, but it is a guideline. It is not a rule. It’s not a regulation. There’s nothing that’s going to require an organisation to do this or behave in this way. Right now as an AI vendor, someone whose company makes AI. There’s no rules, there’s no laws that govern how we do it. There is nobody who checks our data and nobody who checks our algorithms, there’s no certification, there’s nothing. So we, as the entrepreneurial community, estimate that we’re about five years ahead of any rule or regulation that’s relevant to where we are. If we rely on government to be regulating this field, by the time they catch up we’ll be another five years ahead anyway.

I feel now with emerging tech, and this is probably not just AI, although it’s the fastest, biggest sector, it will also be crypto currency, cyber security, blockchain, internet of things. Also very advanced technologies where the law and regulations are well behind. Will government lead us through this to make sure we have good ethical AI? Probably not. They’ll give it a good try, but they don’t really have teeth yet. Will it be the general community and public? Well they know very little about it. It really comes to business leaders and I don’t even think business executives, I think it’s senior management, middle management, good leaders, informal opinion leaders within business, who recognise that we must do this or else things could be quite dire. 

What might be cool, is if we maybe get to illustrate what bias AI is to have your listeners understand it. There’s a couple of cool things that I encourage listeners to do. One of them is to Google the term unprofessional hairstyles and hit images. Then just scroll down through and have a look what’s there. Then do professional hairstyles, hit images, scroll down. You could do best CEO’s in the world, most intelligent people in the world, professor’s style is another good one because I’ve got an academic background. I like to see how Google actually presents itself as what professors are. Going through this, and I’ll let your audience do that because it’s kind of cool and scary, what we learn is that these machines, they’re already racist misogynists. It really is disturbing what you find that the machines are classifying like this. There’s some simple examples of bias in AI.

There’s other examples where algorithms and AI is used by judges in court to determine the chance of an offender re-offending. When they come in front of the judge, the algorithm provides them a score and then that person is accordingly sentenced. What those algorithms were trained on is historical crime data, which had a much higher skew to men of colour. Whenever a man of colour came in front of a judge, he would automatically, the judge would be given by the AI, a much higher score, the person would be sentenced to a much higher sentence. This stuff is going on every day in our world now without people being aware of it. Is it going to be the law? Very slow. Is it going to be senior business leaders? Also don’t really understand it. It must be we as business leaders starting to think about this.

Julia: What are you particularly optimistic about as you look ahead and say, okay, that’s the challenge, that’s the reality. In terms of fixing that and improving that, what are you really optimistic about?

Catriona: There’s two things I’m optimistic about. Let me start with saying that what I’m optimistic about AI, and we really, most of us AI philosophers and thinkers think 50% of it is going to go really well and 50% of it’s going to go terribly badly. Let me start on the optimistic side, what I think will go really well. 

There are some beautiful AI apps for social causes and I’ll give you some examples. One of the heads of product at Samsung had an autistic teenage boy and he had Samsung fund him to build a virtual assistant for autistic children for when they were going to university. Imagine, if you will, on your phone. You wake up in the morning and there’s a virtual assistant greets you and says, “Hey Julia. Today, these are your lectures, here’s the bus that you need to get on. Make sure you have breakfast before you go. Here’s the professor. Here’s some tips and some good things to say to your professor. Here’s the bus that you’ll be catching on the way home. We’ll let your mother know that you’re on the way home so she knows you’re safe.” You use that app, that AI virtual system to guide yourself through your day at university. How would the parents feel about that app? It’s just so beautiful. How would you as a child with autism feel more confident in how you navigate the day? Fabulous.

There’s another app called Seeing AI, which is an app that narrates the world for people with sight impairment or blind. So they would have a listening device, an earpod perhaps, and as they’re walking the app will describe what they’re walking past. Describe the colours, describe what people are doing, describe the landscape. How beautiful and stunning.

Julia: Amazing and enabling, massively enabling.

Catriona: Completely. There is an AI app for domestic violence that’s disguised as an app for a person who’s suffering domestic violence device and it records and collects data and provides that data to legal support. Whether it’s to the police or legal counsel so that the person who is suffering domestic violence doesn’t have to collect all that data and make phone records etc. A beautiful assistant to someone who may be suffering terribly.

Another gorgeous one is called Child Tracker. This is released in India using facial recognition technology to identify missing children. Within the first three days it was launched, 2 930 children who were missing were identified across India.

Julia: Amazing.

Catriona: How extraordinary. What an extraordinary humanity we’re faced with. If we can have robots and machines help us do those things.

Julia: We think a lot in financial services and obviously about payments, we think about financial inclusion and beginning to think about actually how that unlocks potential for, particularly, thinking about people with disability perhaps and their ability to engage with financial services in a very confident way. It’s just incredibly enabling to be able to do that, which is extraordinary.

Thank you for the insights around AI. It is interesting. There’s this conversation around governance and data and how that’s going to change in time and the role of business in terms of really being aware of the biases that exist within it. I don’t want to move naturally on the word bias. However, you were one of two organisations that are female led that is listed on the Australian stock exchange. That must have been a fascinating journey to capital raising, getting listed and all the advisors that come around you and looking at the world of capital markets. I’d love to hear some of your stories from that journey.

Catriona: At the time we listed, which was November, 2016 I had a female chair, Kathy Reed, who’s this most incredible business woman. Much to our great surprise, we learned that we were the second only woman led business to list on the ASX, which is staggeringly disturbing actually. But now that I’ve been in the capital markets, I can see why. It’s what I regard as a hyper masculine environment. I see very little diversity in the capital markets. It’s predominantly males, middle-aged males, predominately white males who are the fund managers, investment managers, stock brokers, etc. It is, I think, quite a journey to actually have investors understand how we as women run businesses and how we might do it a little bit differently. How we present, which might be a bit differently. And of course we know that it’s enormously difficult for women historically to raise capital. Less than 2% of all venture capital globally goes to women led businesses. We’ve actually been quite effective on raising capital on the public market, but it doesn’t come without its great difficulties and indeed I have some great stories.

Julia: I was going to say, could you share some war stories? Could you talk us through a few?

Catriona: The most interesting one was very early in the piece when we, Kathy and I went and we did some roadshow and we were down in Melbourne and we were presenting to a group of investors and we were on our way back, flying back to Sydney. One of the directors got a call and he said, “Oh wow, someone wants to invest $1 million in the business.” And we said, “Great, that’s fantastic.”

Julia: That’s a good call to have.

Catriona: Good call to have, we’re very happy, we’ve obviously done very well. Then the board member, I could just see his blood draining out of his face. He said, “Ah, okay, there’s a condition,” I said, “Of course, what is it? Revenue or reducing the burn or reducing cost of customer acquisition, what might it be?” He said, “No, no, none of those things.” I said, “Okay, well, what else is it?” He said, “Well, this investor will invest $1 million into the business if the female CEO takes her nose ring out.” $1 million into the business, if the female CEO removes her nose ring, that will warrant that little investment.

Julia: Did they become an investor? I think I know the answer to this question before we begin.

Catriona: No. In fact I had a smaller nose ring. I went the next day and got a really big one.

Julia: Imagine if you had said yes. The implications of having somebody have that degree of control and command over something and then what kind of investor they would end up being as well. Incredible. I hear a lot of these stories. In fact, in the UK, Alison Rose, who’s now the CEO of RBS commissioned the Rose Review, which looked at the amount of capital that goes into female run organisations. As a female entrepreneur myself, I understand that entirely. I wonder whether again it is the world shifting. I know you spent time with other women led businesses and mentorship and helping others to move on that journey.

Catriona: It is shifting but in the most marginal way. Not anywhere near the speed it should or what we predicted. A very wise feminist in Australia quoted that it never really gets over 17% of anything. Women, we just don’t get above 17% on boards. We don’t get anywhere near 17% of capital. We don’t get to be 17% of the jobs. We’re just not seeing the numbers move. Now, certainly in Australia, the number of women on boards is increasing and that has had some movement. The number of women led businesses is increasing. The capital going to women marginally increases. I think it’s moved to just below 3% now, but we’re just not winning the goals that we expected to by now. It’s still a very tough road.

Julia: What do you think will drive change?

Catriona: There needs to be, I talk about a shift towards women of power and the good men. You noted in my introduction that I was a woman of influence. When I got this acknowledgement, I went and looked at what are the men’s awards? It’s a women of influence and the men of power. I went, that’s interesting. What does influence mean? Influence means the ability to indirectly affect change and power is the ability to directly affect change. Why can I only influence and the male counterpart can directly change things. That’s not okay. I’m all around starting to champion women into positions of power where they are absolutely able to make the calls and direct change. To support that movement, we need what we call the good men. The good men are those men who recognise that diversity inclusion is an essential means and path to a healthy humanity.

I’ve already quoted today about some of the things that are going to go very badly. Imagine who’s designing the autonomous AI driven weapons at the moment. Those sorts of things are very frightening. We need men who can do a couple of things in the workplace and in community. That is that they can hold space for women. When there’s a woman in a room, if a woman’s trying to speak, they hold the space, so the woman is able to speak in her way and she’s not corrected, she’s not shut down, she’s not mansplained to, that they can hold their space. That they may indeed amplify what she’s saying by either encouraging her just to speak more or picking up a point that she’s made and then making it again. We need men who are supporting and championing women into these roles.

We need men to be calling out the biases and the unfairness that they see. We also need men to understand themselves. In a primal sense, there’s two things that go on for men and women. Women have an innate fear that men will hurt or kill them. Men have an innate fear that women will ridicule or laugh at them. And so these primal things that are at play during performance reviews, during is this woman more powerful than me? Is this woman go to laugh at me? Is this woman going to be promoted ahead of me? All of these things where there’s a sense of threat. Men just need to be aware of what comes up in them from their primal self or the conditioned self and be able to deal with those and to start to really see women in their true right and for the true value that they bring to a workplace. That we have a rightful place at the table in positions of power.

Julia: If we think back to the conversation we were having on stage yesterday, one of the things that really struck me was, in addition to the relationship with men, the good men who are calling out, and they go through their own career paths, their own career journeys. One of the things we’re thinking about a lot on the podcast at the moment is that that hiring manager management level is splitting out into those who are completely agnostic and actually still don’t get the diversity joke, in terms of high performing teams. But that is shifting, is my opinion. The second is those who feel very threatened by that and feel blamed for it, and those who are very positively engaging exactly as you’ve described. In order to address the group that feels most threatened, is the way in which female career journeys and male career journeys run in parallel very differently, but also begin to intersect as well. I know you’ve got some fascinating thoughts around that female career journey. Which I’d love to hear if you’d just share it with the listeners.

Catriona: The heroine’s journey.

My partner’s name is Dr Arne Rubenstein and he’s world expert on rights of passage for men. My son, Hunter Johnson runs a programme called the Man Cave and his two greatest champions are actually Harry and Meghan. Your royalty.

Catriona: Hunter works in redefining masculinity for teenage boys and what mental health as a result of redefined masculinity. It’s so beautiful, I’m surrounded by men who work in this area of what is the masculine journey? Then I started, encouraged by both of these men, to start to look at what is the heroine’s journey? What is the female’s journey? This is what we’ve learned so far. 

The heroine’s journey starts out, with young women, with something called the separation from the feminine. This is as we are small girls we may like to play with dolls and ponies and fairies and pixies, but very soon, probably mid primary school, we recognise that those things are not really valued by society and the things that our fathers are doing or the boys are doing, regarded more highly or rewarded. The fact that our mothers might be, even if they are working, they’re still doing things at home, carer type duties, and we don’t really recognise that those things are rewarded in society. 

We start this separation from the feminine and we start to identify with the masculine. We then early teenage years start to think, “Oh, what’s my career going to be? What are the goals that I want to achieve? What education do I want? What’s going to pay me well? Do I need a husband? Do I need a family?” We start these type of goals setting and start to achieve those things that we think society regards as successful, which are much more strongly associated with the masculine archetype than the feminine archetype.

We go separation with the feminine, identification with the masculine. Then we get to a certain time, and this is very interesting because I think it also relates to that time we know that women start to peel off from the workforce in leadership positions in their thirties and forties. Now it’s a time called the illusion of success. We as women have got to a certain level. We’ve got the house, we’ve got the car, we might have children, we’ve got the job, we’ve got the degree, and we, in all intents, are successful. However, we start to realise that there’s something missing, that that success doesn’t feel as good as we thought. It’s hollow. 

The phase next on the heroine’s journey is what we call this spiritual aridity. We start going, “Okay, I just don’t feel this. I’m not really finding meaning or purpose in what I’m doing.”

The next step from there is what we call the descent to the goddess, which is actually tracking back to our true feminine womanhood. That might be some women take themselves off on a retreat. Some women find themselves sitting around with candles and aromatherapy going, “Who am I and how did I get here and why am I not fulfilled or happy? What is it that I’m missing? Because I’ve ticked all the boxes.” But what on the heroine’s journey has been missing is that it’s a complete separation from the feminine archetype. 

The next phase after descent to the goddess is yearning for the feminine. That’s where you see these big groups of women, networking groups. You see the women wanting to go on retreats and holidays together, and women start to yearn for the sisterhood and to connect back to their womanhood.

The next phase from that is what we call the repairing or healing of the mother daughter split. That’s not necessarily your own biological mother or your own biological daughter, but it’s the healing of your original womanhood, and the daughter that separated and went off to do the masculine journey comes back and goes, actually as a woman now I want to be whole. I integrate the mother daughter, I heal that and I value things that are feminine. I value things that are womanly. The final bit is in the integration of self. So in a dualistic way, we say that’s the integration of the masculine and feminine. I’m moving a bit away from everything, having duality, being masculine, feminine. I think there’s probably shades that are not masculine and feminine, but it is that integration of self. By the time we’ve gone through the full cycle of the heroine’s journey, we are strong, we are whole, we are integrated and we are no longer splitting off into the masculine and the feminine. We’re no longer denying the feminine self. It is there, strongly represented in who we are.

Julia: It’s very interesting when you think about personal identity and when you think about LGBTQI inclusion and people on their journey of finding themselves through how they identify, whether it’s more masculine or more feminine. That’s what really caught my attention when you were first talking through that journey, is that it’s not a binary, it’s partly the male journey, it’s partly masculine journey, partly feminine journey. But actually the two of them come down to the whole and the individual themself. In a different lens from the way in which we have been brought up and taught historically.

Catriona: I think if we look at the gender spectrum, there’ll be plenty of men listening to what we just talked about who’ve said, “Actually I can relate to all of that myself.” Or any gender. We talk about masculine and feminine, but I think it is beyond that and whether it’s the heroine or hero’s journey or any other type of journey, it’s a cycle. You saw how the women that we spoke to reacted yesterday when they start to identify, “Oh, I’m not alone. The fact that I’ve just felt these things doesn’t mean that I’m failing or that I’m the only one feeling this.” In fact, we probably had the whole room who could relate to some of those.

Julia: What was fascinating afterwards, because it wasn’t just an all female audience as well, was actually some fascinating conversations with the men in the room as well who, beginning in part, the first time they’d heard the conversation, but in parts to begin to resonate, some of that begins to resonate with them. Which is amazing. 

It’s been the most incredible conversation. I can’t believe how time disappears on these podcasts. I suppose my next question, my final question really is, as we sit here in Sydney, and it is really interesting, this is being recorded in November. I flew in a week ago with news about bushfires all over the press. I was getting WhatsApp messages from home saying, “What’s it like? What’s it like?” And as we sit here today, it is a smoky day in Sydney. There’s no other way to describe it. I was thinking this morning as I left where I was staying, I’m thinking that there is smoke in the air, and maybe it wafts in, it wafts out. But it’s most definitely here. We think about artificial intelligence, we think about the role of technology and of course the reality of where we are today. And I’m really keen to hear your thoughts on what comes next and what are the big thoughts in your head at the moment?

Catriona: If we think about the three major world challenges that there are today, the things that are likely to threaten humanity. One is nuclear war and we’ve never been at a closer time in history to nuclear war than we are today. Do we talk about this? Not really. Should we? Absolutely. Number one, nuclear war. Number two is climate change and we have pretty good dialogue about climate change. The third one is the arrival of disruptive technologies. These are the three things that are likely to destroy the world. I have had an experience recently where we have a family property in New South Wales, a 10,000 acre farm that we’ve had in the family for over 50 years. And right now 90% of that farm has been destroyed with the fires. In fact the fires now there’s more than 2 million hectares of fires burning in New South Wales, much greater than the Amazon jungle that recently burnt and the whole world was up in arms around that.

Our fires are far, far greater. Fire season is coming a month earlier. We’ve had the same level of intensity over the last three years fire seasons all within a month. It is a disaster. And so my personal experience with this now, with being up at the farm in front of the flames, seeing the animals that are destroyed, the native animals, the cattle, the shelters that have been destroyed, has really got me thinking profoundly about we must put two of those world challenges together and we must start to get emerging technologies dedicated to climate change and to reducing the effects or mitigating the effects of climate change. I first hand experienced this and it was like being in an apocalypse. It was a deathly feeling that the earth was dying and that’s not okay.

Julia: No, absolutely. It has been the most incredible conversation. I’ve had the privilege of spending some time with you while I’ve been here in Sydney, recording in November, 2019 and to have this conversation in such an intimate setting in the boardroom of your office with my iPad recording, but also on stage in front of hundreds of guests at the conference yesterday. I can never tire of hearing you talk with such passion about technology, about talent, about business models and then also thinking very proudly about the future. Catriona, thank you so much for being on the show.

Catriona: Thank you.

Kieron: This episode of the DiverCity Podcast was produced by me, Kieron Yates, on behalf of Julia Streets Productions. You can find out more about the guests on this week’s show on our website, divercitypodcast.com. Whilst you are there, you can also sign up to our newsletter for all our latest updates.

To be sure of catching all our future podcasts, subscribe to our feed on iTunes, or your favourite podcast app. And, if you’ve enjoyed this episode DiverCity Podcast, remember to give us a rating or review. This all helps promote the show to a wider audience. Finally, our Twitter handle is @divercitypod. Thanks for listening.