 Well, hello, everybody. I know Tracey and I are both very excited to be here to discuss this really important topic. Just to set a bit of context for the conversation, I think we both wanted to really make this conversation about the future, about the opportunity, about the solutions. I think everybody here attending this session will recognize that there is a problem around diversity and tech, but we want to look at what's coming ahead. But to set the context, I guess I really wanted to ask you, Tracey, first of all, what does diversity mean to you when we're talking about it in technology? It's difficult to define exactly what I think of as diversity. But to me, it means a lot of things beyond gender diversity, which is what people's brains often go to. There's racial diversity, diversity of life experiences. We're looking at things like disability status, family status, all sorts of different layers of identity and intersectionality. And I think it's important when we talk about diversity, that we're thinking not just diversity, but also inclusion. So it's not just bringing people into the room, but actually setting them up to succeed as well. So I should remind you all that there's the opportunity to ask questions. I have this wonderful iPad in front of me. So please do feel free to ask any questions of either of us and we'll pick them up as we go through the session. One of the things that I found really fascinating when Tracey and I were talking about the issue of diversity in tech was the work that Tracey did back in 2013, I think it was, when she identified that one of the challenges around talking about diversity was that we really didn't have the statistics, the data, to talk about the issue with any degree of credibility. And I guess the interesting factor for me there was to drill down on that a bit and maybe think about what are the other types of technology, how can we think about technology as a tool for understanding diversity better? I think the first stage, as you've already mentioned, is getting that data. And where we were in 2013 was that there was no diversity data at all. So in terms of managing the problem we had nothing to work with. Now we have slightly better data on demographics. I think we can push more in the data in terms of getting more fine-grained and looking at intersectionality. If you look at most of the diversity data reports coming out now, they'll have gender splits and racial splits, and then they look at engineering, non-engineering leadership. But it's hard to see, for example, how many black women do we have? And at different layers of leadership, what does that look like and what does the progression look like? And so if you just look at the breakdowns, the snapshot in time, we're not understanding retention issues necessarily. So I think we can do better just on gathering more data and building out those time patterns. I also think we can use technology in much more interesting ways in the hiring process. So there's a number of startups out there that are trying to address bias in the recruiting pipelines and how job descriptions are written, as well as in actually managing companies. So if we look at actually understanding inclusion in companies, engagement surveys are very useful. So understanding that layer of data, not just the demographic splits, but how are people feeling within their companies? That's very valuable information. And general introducing more standard processes around retention and management of people and promotion will give us a lot more opportunity to move forward. And to just drill down on that a little bit, one of the things that people talk about a lot is the risks around things like artificial intelligence because it tends to be designed for particular demographics. How do you think we can try and steer that towards a more positive use of artificial intelligence? Right, so AI is interesting in that there are many different ways for this bias to come in. One is in the models that are built themselves. And so if we have people who are building models that are inherently bias, no manner of training data or adjustments will fix that. You can also bias in the data sets themselves. It's useful for the AI practitioners who are building these systems to have an understanding of the domain and know what to look for. So I don't think there's any blanket solutions. I think some of it is just deeper understanding of domains and knowing what characteristics are being used and how they might be biased and going into do gut checks. One thing that has been a little bit concerning is the use of AI models as black boxes and not inspecting what's happening inside. And some of it is because some of these models are not interpretable. That's one of the dangers of people I've said about neuralness, that they're not understandable. And so it's very easy for bias to just be baked in and we accept it for what it is. I am starting to see some work being done on interpretability of these AI models, which would be helpful. So I think continuing to push on that and making sure that when we're using technology and AI that we understand what we're doing with it is going to be very important. But there's no easy solutions to this. It's really, again, understanding the domain and just being attentive to potential issues when we're plugging these systems in when knowing what the outputs are and what we're going to be using them for. So I wanted to turn to Project Include, which is something that you co-founded alongside a number of other female entrepreneurs, engineers in the tech field. And really think about what are the tenants that you want people to be thinking about? What are the value sets that you want people to be thinking about as a result of engagement with things like Project Include? What do you think are the issues for all of the entrepreneurs, the companies in the room? What should be on their minds? Yeah. So you asked that question very well. Usually what people will ask is, what are the top three things we can do to fix diversity at my company, which is not the right question. Your question is much better. We're thinking about values in inclusion and intersectionality. So thinking about diversity beyond just gender diversity and really these intersections of identities and experiences and making sure that when we're solving for diversity, we're doing it in a very holistic way and not just broadening our circles of exclusion. It's important that we are thinking about comprehensive solutions that are not one-off strategies. A lot of times people will grasp onto easy ideas like anonymizing resumes or applying the Rooney Rule and they want to be able to just check those off and say that they're done. But as with any business strategy, there has to be something at a higher level before you start plugging in the tactics. And another important piece of this is having metrics to understand progress, be able to set targets and have accountability. And so a lot of the companies now that are starting to get into this are still in the phase of trying out different strategy or trying out different tactics. It's important that we start to think more holistically and actually measuring what is successful or not and then being able to experiment with different things and learn from what has worked and hasn't worked. So I'm going to go straight to one of the questions that have come in and please do send some more in. We'd love to try and address them in the conversation. And the question that's come in is really to ask us to share some examples of positive change that we've seen over the past year. Is there something that springs to mind? I've got one, but I'll let you go first. I think the increased awareness around diversity is a good thing and increased awareness of the dangers of the lack of diversity. I would like to hear what your positive observation is. It's probably linked to yours seamlessly. But for me, I think it's about that everybody's recognizing that whether you are a man, a woman, whether you're gay or straight, whether you're BME, whether you're white, you can be part of the solution as well. So I think it's about a recognition that everybody needs to think about this and do something about it. Yes. So to move on to another theme, I guess one of the things that is often talked about is 85% of consumer purchasing decisions are made by women. And there's a whole range of technologies that we're missing out on because women or other underrepresented groups aren't part of the design process. Tell me a bit about what technologies you think we're missing out on because of that. Oh, there are so many. But yeah, that observation that people tend to build products for themselves is a very accurate one. And it makes sense that that is the case. It's very easy to design a product for yourself. Your customer feedback loop is asking yourself if you would use something and if it works for you. So it's very natural that that happens. But then the problem then, without having more diversity in our tech workforce, that we're missing out on so many things, so things that are more targeted at women and some of the perspectives that women will bring, underrepresented minorities, looking at different socioeconomic classes, and the impact on those, so some of the examples of places where we might be missing out, women's health care. One example from a few years ago when Apple launched iOS 8, it had health kit in there for all the different things that you might want to track about your body and your health. And they missed period tracking, which seems like a very obvious, quantified self type of thing that women would use. So just a very obvious oversight, I think. We're talking about some AI models that are biased. There have been examples of Google Photos. Black people were getting tagged as gorillas. There's just really painful mistakes like that. And there's also more insidious things with tools that are being developed in criminal justice. So we're not necessarily serving the needs of people who would benefit the most from technology, and instead just reinforcing biases that exist already. So I started off this conversation talking about opportunity. And I wanted to turn the conversation to one of the real areas of opportunity that's been proven with a lot of the studies out there that actually you get better productivity, better products, better outcomes for your company if you have more diverse teams. Tell us a bit about your experience in this space. Yeah, so there is a lot of research in this, decades of research, in fact, on diversity making teams more creative, more productive in the innovation context. So it is important that we are looking at the innovation context, but being in tech that is specifically the context we're in. And I think what's happening there is when you have teams that are more diverse, you no longer lapse into assumptions that everybody else thinks the same way that you do and you work a little bit harder to justify your train of thought and your perspective. So this sort of finding around diversity, improving the outcomes of teams has been shown even with people of political diversity. So putting Republicans and Democrats in a room and having them work on problem solving, they're more effective because they no longer assume that everyone thinks exactly the same way that they do. So that, she's a very natural thing, even absent the end product. So it's pretty obvious when we're building consumer products that are for mass market, it's good to have representation in those teams. But even just in the innovation context, generally you're trying to solve a hard enterprise technology problem. Having diversity on your team will make you more productive. And how does that conversation pan out for investors? I'm imagining there are a few investors in the room today. What should they be thinking about when looking at diversity? Well, our investor pools, our VC firms need to be more diverse as well. I think the stats are pretty grim if we look at percentage of women and minorities in investment decision making roles. 13% in Europe, according to the State of the European Tech Report that Atomica released earlier today. Only 13% of VCs. Okay, yeah, and it's also lower when you go to bigger firms. So some of those numbers are being padded by women who are striking out starting their own seed funds, for example. Which is good. But I would also like to see the big firms that have billions of dollars of assets under management to be including more diversity in their decision making ranks. And the percentage of capital that goes to women is even less. So even if there's a certain number of female founders being backed, they're not raising those mega rounds that we're hearing about. And so what's happening a lot of times, especially early stages, when the investors are looking at opportunities to put in money, it's a lot on gut feel. And if they are driving with the founders and feel like they believe that this founder will be able to take something and make it big. And the lack of diversity amongst those ranks of investors means that they're missing out on a lot of interesting ideas, which are targeting markets that they're not as personally familiar with, or founders that don't remind them of themselves or don't trigger the same sort of pattern matching that they're so used to. And that pans out into economics as well, doesn't it? Because I think the stats show that investments in more diverse C teams result in about a third more profitability for those companies. So there's a pure economic driver, I think that investors sometimes miss by following some of the patterns that you suggest. That is a great return for their LPs. Yeah. Well, the one interesting point there, it's sometimes hard to say exactly what is at play. It may be that women have a harder time raising money and so having a woman on the team means that they have to try harder and so they have to work harder to justify an investment. So maybe the filter is harder for them to get through. I've noticed this a bit in engineering where there are also not very many women. The women that stick around for a long time have to really want to be doing it because there's so many other headwinds that they're fighting. So I'm sure that there are many different factors at play, but that may be one of them. So I wanted to turn to some other questions that have come in and the first question was just, what would you recommend doing to combat the imposter syndrome? I actually feel like after Trump got elected, nobody should have imposter syndrome anymore because you clearly don't need to be qualified for one of the most important jobs in the world. I think being surrounded by good people that can affirm your abilities and experiences and potential is very helpful. So much of this is a human problem and I found it to be very beneficial to myself personally to have friends who are just there for me. I think it's good for managers and leaders to reaffirm the abilities of people who are from these different backgrounds and to be sponsors for them and pull them up. And it's very helpful when people higher up are saying, I believe in you and really helping putting their name out there and pulling them up. Yeah, I would absolutely back up the point about people around you. I have an incredible tribe of female founders. We have a WhatsApp group. We talk about everything from investors to birthday cakes to the most depressing thing that happened to you that day. It's a very powerful network. One last question, because I'm conscious of running out of time, just to ask about the theme of can tech companies leverage machine learning and AI to pinpoint unconscious bias in the workplace? What do you think about that? I think it's possible AI and ML are tools that we can use but they're not going to be the solution to everything. The people who are designing these tools have to understand the domain that they're working in and know what they're looking for. Fantastic. Well, I'm afraid we're running out of time so I just wanted to say thank you so much, Tracy. Thank you so much for your... It's been a pleasure talking to you and thanks everyone for listening. Thank you all.