Zencastr
00:00:00
00:00:01
Speed1x
Format▸
Share
Embed
Report

E37 - Will AI make the gender wealth gap worse?

Women Changing Finance
Women Changing Finance

0 plays · Nov 18, 2026

Transcript

Speaker: Hi, welcome to the podcast Women Changing Finance, where you will discover amazing women from all around the world who are making finance become a force for good.

Speaker: Welcome to the podcast Women Changing Finance. I'm Christina Tora, and I'm welcoming today Mara Bollis, the founder and CEO of First Prompt, an initiative that helps people build confidence and practical skills with generative AI through hands-on peer-based learning.

Speaker: I have known Mara for a while and have always been very impressed by her passion and her energy to what's right and to go beyond what is expected to create change.

Speaker: Mara, I am so happy to have you on the show today. So welcome. I'm so happy to be here. I'm delighted. Thank you for having me. I want to ask you something quite simple.

Speaker: Your career has spent international development, development finance, traditional finance. What you do today may seem quite different. What sparked your interest in generative ai and why did you feel it was important to create FirstPrompt?

Speaker: Tell us about this journey. Well, thank you so much for that question. And again, thank you for having me on. I have always been really focused on overcoming opportunity gaps.

Speaker: I think my passion behind that is the fact that I think that there's so much human capability that gets overlooked because people don't look the part, they don't have the right accent, they're not the right gender, they're not the right race. And that I find infuriating, honestly, or they're not from the right country.

Speaker: And so overcoming those opportunity gaps has really been the mission of my life. It started in international development finance and then moved into the impact investing space where you and I met.

Speaker: And that then merged into a focus on gender lens investing when I became kind of more sensitized to gender disparities that I wasn't as aware of before when I was strictly in a conventional financial role. and it's been so interesting to me to look at how women see the financial industry different differently than men do.

Speaker: And I've studied that quite a bit. I've studied women's behaviors in the space, the demand side considerations, the sp supply side considerations. And I was doing, i was doing a fellowship at Harvard Kennedy School while ai policy questions were exploding.

Speaker: And I saw so many similarities in terms of the concerns and the way that women's voices weren't being factored in the conversation.

Speaker: Data wasn't necessarily being gathered about women's unique perspectives. And I was shocked by this. And so I went on a mission ah to find the research and I found seminal research by a number of professors, but ah particularly a professor at Harvard Business School, whose name is Ram Koenig, that found that there is a gender gap in generative AI adoption of about 22% and it persists globally.

Speaker: And so, of course, having gone through this experience with gender lens investing and financial questions, my thought went to how are women likably to be harmed by that?

Speaker: And how can women's perspectives strengthen the overall industry by having their voices heard more? And then i just became so taken by that duality in the AI space of it being a set of tools that can garner considerable concern, but is also really necessary for women to adopt in order to further their careers and the careers of others. And so how are we going to navigate that reality together? Wow, fantastic. And it seems such a an important and yet completely unique perspective.

Speaker: on AI. So tell us more about First Prompt. What's the mission? How does it work? And if you can illustrate with some examples. Yeah, yeah, yeah. Thank you so much.

Speaker: So the mission of ai first prompt is really to overcome this opportunity gap in generative And my reflection, other than kind of the research on the gender gap that I mentioned, which that's not the only study, but that's a meta-analysis, so it's very comprehensive.

Speaker: Other than that, I also had my own experience in terms of learning from the available resources that exist on how to get up to speed on AI. And I felt that they were sort of foreign and mechanical.

Speaker: right ah A, a computer scientist talking to me over a hour or webinar that I don't have time to watch and in a way that feels very abstract when actually an abstract and esoteric in a way that's not necessary in the space because these are generalist tools.

Speaker: But that doesn't mean that you don't need instruction in order to learn them. But it's different than consumers, than computer finance learning. You need to know some of that in order to understand how the systems work.

Speaker: But then in order to like deploy them for what's best for your life, really requires more of a hands-on approach, a collaborative approach, peer learning.

Speaker: And Christina, this is exactly what I saw when I was working with women's savings groups. was what is so empowering is women and men working together to sort through the complications of a tool or an approach that can seem foreign.

Speaker: If you have an expert coming in there standing at the front and being like, well, let me tell you about all of the complications and the technical pieces of finance. You get this resistance where it's like you throw out technical terms, it becomes inaccessible, people back off.

Speaker: And there are real parallels here in terms of learning AI. And so I wanted to create ah really peer powered spaces where people could admit, I don't know what the best tool is.

Speaker: I don't know how to prompt. What is best practices? What is a prompt? Like, let's take all of the anxiety down a lot, right? and have a conversation between intelligent peers who have their own competencies in their respective fields of interest and are learning how to adapt this tool on their own terms.

Speaker: And what happens is once you take that pressure off and provide people with some basic rules of the road, they get their power back.

Speaker: Right. It goes from being this kind of intimidating, overwhelming thing that's taking over the world to a tool that I can use to empower myself and empower others. Now you need to know the guardrails stepping into the space. That is absolutely true.

Speaker: But I think I set up first prompt because I think there needs to be a different type of learning approach than what, than what I was, I saw that was available. It's, I think that that's really important, as you say, is both very hands-on and at the same time with the guardrails, with an ethical approach, very, very important.

Speaker: So I understand it's quite a new organization that you set up, but you have examples of rolling out some workshops or experiences for for learning.

Speaker: I do. My gosh, this is the best conversation ever. Yes, I do. I have stories that I can share. So I was working, one example, i was working with a women's empowerment social enterprise in Africa.

Speaker: And I met the one of the founders. And in our conversation about kind of walking him through the basics of of good generative AI use,

Speaker: he realized that he wasn't sure. He thought that there might be disparities in his own organization with respect to men and women's usage of generated AI.

Speaker: And he said, as someone who runs a women's empowerment organization, i can't be perpetuating Right. Inequalities in my own team, that's not acceptable. And so we started to work together in order to try to close those knowledge gaps and comfort gaps.

Speaker: And the first thing that we did was ah like an intake, right, to understand the organization, where they are, etc. And then we did an an anonymous staff survey so we could see the disparity of usage.

Speaker: And we found, and this is very common, that there was a lot of usage of private chat GPT accounts that were free accounts.

Speaker: Because there's such limited instruction, the first thing that we needed to do was to clarify to everybody on the team, you're invited to experiments. But you need to do so with tools that are safe from an enterprise standpoint.

Speaker: So the first thing was to migrate them from their private chat GPT tools to enterprise Gemini, right? That was already a win. Like we could have stopped there and just celebrated. because Now there's encouragement and messaging from management that we want you to be experimenting.

Speaker: Please be trying the tools, right? Part of the reason women hold back is because they don't know the rules and they think they're going to get in trouble. So once management is messaging, we want you to use these to maximize yourself and others. It equalizes the gender disparities very, very quickly.

Speaker: So that was the first piece. And then the second piece was now we had a better control of the data privacy issues because we were able to migrate the team onto a secure LLM that could protect their data.

Speaker: But we didn't stop there. We did the survey. We came up with a policy. And then we did a training that was remote, three hours, very hands-on. And it was a mix of kind of general rules of the road, clarity on guardrails, being very expressive of what the policy is and how it should be. and we had a big Q&A about to socialize their policy.

Speaker: And then we broke out into... breakout groups that were function specific with function specific exercises. And what was so interesting after it was the reflection after that conversation, after that workshop.

Speaker: And one person said, before this workshop, I thought everything that chat that AI produced was gospel. And now I know there are hallucinations. Now I know need I need to check the work, etc.

Speaker: And then another person said, before this workshop, I thought everything AI produced was crap. But now I know that there's this, that you need to set the context, you need to be clear about the task.

Speaker: you need to provide it enough instruction and guidance in order to produce something useful. And that is the arc, I think, is very common. And those reflections are very common. And it's exactly why this type of work is needed.

Speaker: In the process, we also had a number of laughs. Like, I want these spaces to be joyful because learning something new that challenges your expertise is tough.

Speaker: It's not easy. I'm very sympathetic to that. And I think if you leave people feeling better than when they came, there's a higher likelihood that they'll continue the behaviors that you introduced within that setting.

Speaker: Any funny moments that you want to share? Yeah, there was. Well, there's one funny one where there was a team member who we were doing the reflections and there was a team member who was working out of a refugee camp at the moment.

Speaker: And she's giving a very impassioned reflection about her feelings of AI. And then all of a sudden there's a helicopter that's coming through and she's like, Oh, sorry, there's a helicopter next to me. And then the helicopter passes. And then without a beat, the woman is just so on top of it and she's so brilliant that she's like, okay, the helicopter's gone, I'm going to keep going.

Speaker: And so I was like, that's amazing. Someone who can keep their train of thought but while doing a training in a refugee camp, while a helicopter is landing next to her, and then not without missing a deep beat, just kept on with her intelligent reflection. It's like, that is a champion. What an amazing woman.

Speaker: That was an inspirational and funny one. We all laughed. But then there's another one where for anyone who's ever done an AI training, you set up demos, live demos, and 50% of the time ah they work.

Speaker: and because the Instruments don't necessarily replicate their behavior predictably. And so here I am rolling out this exciting way that they can do quick and easy data analysis using Gemini. And I've already shown the result, but I want to show them the process.

Speaker: And it completely fails and doesn't know how to do it. It's like, oh, please don't lose faith in this training. that's These are actually really powerful tools, but this is also proving the point that you cannot trust them completely.

Speaker: That's funny. More seriously, I wanted to ask you because at some point in your example, you explained that you moved people from Chagipichi to Gemini. Is this an overall recommendation? Is this something that in this specific case was required?

Speaker: i am so happy you asked this question because we can do some sort of um guardrail setting right here on this call. A lot of people don't know that kind of like social media with generative AI tools, we, the users, are the product, not the customer.

Speaker: By which I mean, if you are using a free tool, that does not have privacy settings, what you are paying for that service with is by providing your data.

Speaker: And that data is getting hoovered up into whatever generative AI tool of choice and being used to train the model, right? This has real privacy and data privacy risks.

Speaker: And I can talk about that separately if you want, but believe that to be true in terms of you don't know how your data is being used, how it's being released, to whom, et cetera, once that's been hoovered up, yeah? In chat GPT, in order to have privacy settings turned on, you must have a paid version.

Speaker: If you are the head of an organization listening to this right now, if you do not know, your team's usage footprint, there's a very high likelihood that your team is using a generative AI tool without privacy settings.

Speaker: turned on, in which case you do not know what data is being put into those systems. And there's an organizational risk that's being created that you're not managing.

Speaker: So my recommendation is to, the best case scenario is to migrate everybody onto to an enterprise solution, like we did in this case with Gemini, because they owned the Google Suite,

Speaker: as a social enterprise and everything, all data that was entered into Gemini, which is Google's AI tool, is held secure just as if you were putting a file into ah Google Drive.

Speaker: It's the same level of protection. It's the same thing with Copilot, right? mean, if you're using Microsoft Copilot in an enterprise setting. So that's the best case scenario. The not great, but okay case scenario is that If people are going to use their own tools, and this is a longer conversation in terms of how you parse these things, but at a minimum, if they're using their own private tools, they must have privacy settings turned on.

Speaker: And that in ChatGPT is only ah possible if you have a paid tool. In the cloud, which I love, it's possible to turn on privacy settings, even on the free version.

Speaker: And it's also considered a more ethical tool. Great. Amazing insights into AI and what to do. Super. Now, I want to understand also another kind of part of your work, which is, I mean, not another, but the core part your work, which is about peer-based and community learning.

Speaker: So why is that approach of community, of peer-based, so critical for ai adoption and the work they could do? Well, so it's a couple of things. One thing is that the group of people that I'm targeting with my work are mature adults.

Speaker: And best practice in terms of adult learning is a peer-based approach. But lectures don't land in the same way for adults, and particularly adults with who have a considerable amount of expertise already.

Speaker: You're also... leaving expertise on the table by not taking advantage of what they already know in the learning process. And why I'm focusing on kind of mature adults is because I think that this the gender gap in AI and the opportunity and gap in AI is looks different for younger people because of, I mean, this is depending on the school system, at least in the US, I think a lot of this is going to be leveled out through ai integration into high schools, et cetera.

Speaker: But my real concern are the people who are totally outside of an academic setting and where they're in their respective jobs and people, they're receiving very little training on the job, but they're actually experiencing a lot of pressure to use the tools.

Speaker: So I think it's the most effective approach. And I think that iterative learning and experimentation is what works both for mature adults, but also i think that women prefer a more slot form of interaction that's authentic and where you can build interpersonal connection while you're learning and helping each other. In fact, I'll tell you another thing.

Speaker: So this is a hot take, Christina, but a friend of mine who I met at Kennedy School and are starting a podcast ourselves. It's going to be called Women's Blaining And why we're calling it Women's Blaining is because of this peer element that we think that how women like to learn is in a conversational iterative way where nobody knows the exact right answer, but we're kind of piecing it together together.

Speaker: And so that's the vibe that we're going to be introducing into this podcast is one that is not expert centric, But that is you know, we're all exploring, we're all learning, come learn with us. So if you're thrown in a conversation where people are like, Oh, what do you think about the latest nano banana release?

Speaker: You're not caught off guard, you know that it's happened, you have an opinion, because you will have listened to people who have various opinions. And you feel more empowered in those types of spaces and conversation, you feel like there's a space for you in this conversation.

Speaker: It's amazing. i love this idea. Let me know when it's out. You just talked about a specific example, but what kinds of recommendations would you give to organizational leaders today based on what you know?

Speaker: I think that organizational leaders who don't have visibility into what's called shadow usage. So it's where people are adopting different generative AI tools, but they're doing it privately and individually.

Speaker: I think that if you don't have visibility into that, what you're doing is there are a lot of risks that are produced by the LLMs, right? There's a really well-known case this past summer here in the States where I think it was the Chicago Tribune released a list of the top 20 books of the summer and half of them were hallucinated.

Speaker: So the the person who wrote the article had just gone in, grabbed the results, hadn't checked it. The editorial team apparently didn't check it and they published it. And then authors were getting called and saying, oh, congratulations, I didn't know you had another book out. And they' like, I don't.

Speaker: That's a real organizational risk. They look like morons, right? And so there are these risks that are flowing into organizations downstream from the LLMs. Like, let's talk about bias, right?

Speaker: These LLMs produce very biased results. And sometimes that really bites companies in the ass. There is this well-known case early on where it was found that Apple wallets,

Speaker: were offering credit limits that were much lower for women than for men, which is, I was algorithmically determined. And then it became a huge scandal for the company, right?

Speaker: So where, for organizational leaders, my first thing is Do an anonymous survey of usage within the company so you understand who is using what and for what and why Because, and it can't be punitive.

Speaker: Because people can't be punished for things that they've never had guidance for. It is now unacceptable to not have guidance to give people guardrails so that they know how they can use it and how they shouldn't use it.

Speaker: And then that take that information and build it into a policy that can be socialized. And even if the socializing is in tranches where the first message is okay,

Speaker: no putting our company data into any private tools. Like you have to set up the guard. You can set up very quick and easy guardrails initially, and then go into more substantive policy setting once you have a sense of the usage. And it's important to carry a balanced message, right? Is like, what can you use it for and what you can't you use it for?

Speaker: And like I said before, that in and of itself, will help level gender disparities. Because again, there was this research, it was PhD students somewhere in, I think in the Scandinavian countries where they where they looked at male and female student usage of generative ai And before the courses had a policy, men were much more likely to use it. And once there was a policy set for the class, then usage evened out.

Speaker: Women just really want to know the rules of the road. i mean, they also have additional concerns. But as a first step, you need to know who's doing what and you need to let your your teams know what they're allowed to do and what they're not allowed to do.

Speaker: Great. Okay, that's fantastic advice. i also want to ask you because prior to what you're doing today, you were in impact finance and development finance.

Speaker: How do you connect your work today with your previous experiences? And maybe also, do you have any insights into how what you're doing, how AI or the work that you're doing today can help finance...

Speaker: be used in a better way, in a way that creates change and positive change. So how this relates is, it was kind of a stunning epiphany to me, how we're doing the same thing all over again with ai that what happened with finance.

Speaker: So I'm very inspired, and I think you are too, very inspired by the book Invisible Women that came out several years ago. But in homage to the author, Caroline Corotoperas, I will say I was very influenced by this idea of women um struggling to exist in systems and structures and institutions that they had very little part in creating.

Speaker: And this was completely my experience in interacting with generative AI tools where there are moments, the results that I get feel so foreign and not for me, that is very disconcerting.

Speaker: Typically from an an image bias perspective, but other pieces too. And it can be very disconcerting. I see the same experience in finance all the time. We did a research that was looking, it was participatory research in Guatemala that looked at men and women's impressions of seeking small business finance.

Speaker: And from an attitudinal perspective, there were many times where women respondents reported not feeling welcome in financial institution spaces.

Speaker: There was one particular experience where there was a woman who was waiting in line to speak with a customer service representative and she kept getting, they kept calling the men up in front of her before her in line because they just assumed that she was waiting for someone and they couldn't imagine that she was actually a customer.

Speaker: And with a i I don't think that it's an intentional, I don't think there's intent to exclude. Of course, they're businesses. They want to serve as many people as possible. But because the diversity levels are so low in the AI space, there's just been less opportunity for women to engage and to shape the space.

Speaker: So there are a lot of parallels in terms of that. And i think from a One thing to, one kind of message for the financial community in this is something that we talk about in Gender Lens Investing all the time is that that disparity breeds possibility, right?

Speaker: Because women see finance differently, because women build different types of enterprises, because women are building different types of AI tools, they are seeing gaps in the market that men just won't identify because it's not part of their life experience.

Speaker: So one message that I would have for investors in terms of AI is consider what, look at the founders, right? Women do, they migrate more to creating solutions that are appropriate for, say, the healthcare space than to creating like sex bots.

Speaker: you know essence There's a lot of game creation and there are different value propositions that women in the space are offering that may be overlooked because they're not the loudest or the most visible voices.

Speaker: In terms of how AI can improve finance, I think that it is possible for AI to be a leveling instrument and to offer, ah take some of the subjective decision-making out of questions like hiring and investing, et cetera, but only if you're really intentional about it. Because right now, the way that the algorithms are set up is they're trained on historical data in which biases are encoded.

Speaker: I'll give you an example. This is a crazy example, Christina. So I asked all three, the three of the main ai tools, main generative AI tools, tell me the story of a little boy and a little girl as they grow up and their personal lives and their work.

Speaker: That's it, whole prompt. Every single one of them made the boy grow up to be a man who was in the sciences. Several had gone to MIT. They're running architecture firms or something scientific, something technical where they're earning considerable money.

Speaker: Every single one of them had the woman in educational fields, nonprofit, doing policy work. And then I asked afterwards, what is the gender wealth gap?

Speaker: that will result if this kind of were to scale, these outcomes were to scale. And it's massive, right? So if you were an investor and you were going in and saying, I'm thinking about investing in Shirley, who's running a med tech firm, how much should I value and et cetera, et cetera, how much should value her company?

Speaker: It is very likely that the valuation you get in will change if you adjust the name from Shirley to Peter. So now if if you constrain your data, like in a RAG system or something where you've kind of like, you've put your data in a particular box and you prompt it correctly, there's a possibility that it could level some disparities where maybe a PE and investor would just be looking at the, bot but just feeling the vibes.

Speaker: and wouldn't be looking so much at the data. But for the most part, what I'm seeing in the finance space now is likely to increase inequalities. So what we really want to see is investors and financial institutions that are recognizing this and that they're putting themselves out in front as ethical providers of these services that are aware of these biases and recognize the fact that they could be missing massive opportunities if they're not training their models to overcome these inherent biases.

Speaker: Wow. Fascinating. i hope that you know many of the investors listen to this and take your advice. Because indeed, i mean, we're already in a world where everything is going to grow in the inequalities. so this is an additional one that can be there.

Speaker: How to shift to another question from here? Yeah. I want to go into a more personal question for you because your journey has been impactful, but this is a very impactful job that you have now, even more than a job, it's a mission. This podcast is also an opportunity to reflect on your achievements, on seeing different things that you have done.

Speaker: Tell us about what you're most proud of what you have achieved. Let's celebrate together. Oh, that's such a lovely offer. i will tell you the thing that I'm most proud of, the the outcome that I'm most proud of, came from when I was a fellow at Harvard Kennedy School.

Speaker: And I was teaching a seminar on bringing a gender lens to development policy and practices. in my seminar was the prime minister of Cote d'Ivoire as a student.

Speaker: And he was there not because he was particularly fascinated by gender dynamics, but because someone had, he was going to be teaching a course and someone had said, oh, Mara's taught her course before. You should sit in on Mara's and see how she teaches so you can get a sense.

Speaker: So he comes to the first class and he dutifully sits through class and he's very quiet. and He's at the back of class. And he approaches me afterwards and he says, Mara, I think I should not come back to the next class.

Speaker: And I said, why? And he said, I'm the only man. i think I'm making people uncomfortable. And I said, not only should you come back to the next class, but now you know how the majority of professional women feel all the time.

Speaker: which he could have reacted poorly to because he was a former prime minister. He looked at me and he said, yes, Mara, I do know that now. keep And he came back for every class.

Speaker: That's fantastic. Anyway, fast forward. he goes through the seminar. He's very moved by the data, but also the stories that the students shared in terms of their own personal experiences. And he said, I need to do more in Cote d'Ivoire. So he's still ah a governor of a region Cote d'Ivoire.

Speaker: And so that period of time together at Harvard kicked off an engagement that we, a consultancy and engagement that we had the subsequent summer where we did participatory research.

Speaker: We did multi-stakeholder conversations. I did trainings on the gender gap with legislative and elected officials.

Speaker: officials And we got several new projects launched from that. And the one that i brings me to tears actually came out of a community social worker who was at the participatory at the state multi-stakeholder event.

Speaker: And she pointed out something that all of the experts in the room were missing, which was that we wanted to pursue more work on women's economic empowerment. That's great.

Speaker: Do your entrepreneurship programs, do your financial inclusion programs. But what you're missing is that many, many women in our region don't have IDs because they weren't born in hospitals. Right.

Speaker: And it may not have seen administratively necessary or not, it may not have seen necessary or clear why you would have to go out of your way in order to get an ID for your child, subsequently adult person, woman, who's now not able to get basic benefits.

Speaker: She has no identity. And there was a World Bank project that was starting up. And as a result of our work together, they decided to go into this region as well. They added this region on.

Speaker: And so I'm hopeful that that first conversation that I had with the prime minister where he almost left has led to more women being able to claim their identities in rural Cote d'Ivoire.

Speaker: amazing fantastic story thank you mara for sharing i wish we had more of these stories because it's really transformational you mentioned this a little bit but this is a question i'm asking all my guests how do you feel as a woman doing your job so I feel so grateful.

Speaker: There are a lot of inequalities that women face. We're not nearly in enough leadership positions. There are so many ways in which we still have to advance. We're going back in terms of women's health rights in this country. There are a lot of issues, but we are very good at building connection.

Speaker: I think I'm good at building connection and that's what I try to do through my work. I mean, connection in order to empower, right? My whole thing is you need to learn AI in order to be able to empower yourself and others.

Speaker: You also, we also need to be very cognizant and aware of the risks of AI and hold the companies to a very high standard, right? I... love the vulnerability of being able to say, i don't really know.

Speaker: And the connection that that can build and that I'm able to work in the space of AI, which is so fascinating in order to help other people feel more confident and to build opportunities for people to feel more connected.

Speaker: think there is something that makes that easier for me being a woman than it does for a man. I mean, we can have a long conversation about the stuff that men are up against and it's a lot.

Speaker: So I feel really grateful. It's such a great feeling. We're approaching the end of this conversation. So I have three quick questions for you.

Speaker: And I might even have a bonus question. yeah ah What is keeping you optimistic? What is keeping me optimistic? There is something like keeping me optimistic. I think we're in a moment of time on the gender and kind of overcoming the opportunity gap in AI.

Speaker: where there is the possibility of these very fruitful networked connections across boundaries and across geographies. There was one day where I had conversations with people in Uganda, Turkey, Latvia, England, and it was one day.

Speaker: and I'm sitting in Boston, Massachusetts, and I never left my office. I hadn't left my office to meet these people. We were connecting because we were feeling this that the there's a movement is like so overused, but there was this moment of energy around how to utilize ai to empower ourselves and others and where people want to be sharing their resources and learning from each other. I had a woman who's doing very similar to work what I'm doing in Turkey. And we both opened our books to each other. and we're like, here's how I do my workshops. How do you do it? i mean, no proprietary anything. it was just, I want you to be successful because that will make me feel like I've been more impactful in the world.

Speaker: And it's a beautiful energy right now. I think because the space isn't overcrowded yet. And that the people aren't feeling guarded or like in their little fiefdoms. They're feeling like very free to make connections and build possibilities. So and the energy is amazing.

Speaker: That's fantastic. And I love your example because it' also busts kind of a myth that AI is only for people in developed countries and that a lot of people are missing out or not part of the conversation. And it's,

Speaker: you know, what you're just sharing is not true, right? My God, everyone should check out women, Ugandan women in which is a LinkedIn group. It was one of the first sort of peer conversations that I had.

Speaker: They're magic. They're doing such interesting work. There's pockets of, there's my pockets, but like a quilt that, of such incredible activity in this area that's global and there's a a hunger for people to be connecting and sharing notes and seeing how people can help each other.

Speaker: Fantastic. So second question, what's inspiring you or who is inspiring you? You know who's inspiring me? It's Claudia Golden. Professor Claudia Golden from Harvard who was the first woman to win the Nobel Prize in Economics.

Speaker: First, and this happened, i don't know, three years ago. She's the first woman, first individual woman. There are other women who've won it with men, but she was the first woman to win it on her own.

Speaker: And in this country, in the U.S. s and many other countries, it is becoming increasingly difficult to talk about to work proactively on gender disparities. It's a humble brag, but when I was at Harvard, she Claudia Golden had received the Nobel Prize and was getting a lot of acclaim for her work. And she came to speak at the Kennedy School and I was like, I've got her email address. I'm gonna email her.

Speaker: So I emailed her I was like, hi. I'm Mara. I'm teaching the seminar at Kennedy School. Would you mind if I came into your office hours? And she was like, sure.

Speaker: And I brought my male TA who was helping me teach the gender class in generative AI. He's a massive gender equity advocate. And we went together and I asked her in this, in the context, that in the political context that we're in right now, how do you do work proactively in order to level inequality, particularly gender inequality? And she said, don't call it gender inequality.

Speaker: So do the work, but don't call it gender inequality. I still have a hard time practicing that because I'm not as disciplined as I should be, but it's true. I mean, these issues and in terms of like the opportunity gap in AI, yes.

Speaker: There are women who disproportionately face this, but like the example with organizational leveling of awareness around AI use, that is doing gender work, but I don't go into the conversation going,

Speaker: hey, you know, women are falling behind, you need to level them up. I say there are real problems with shadow usage. You need to be explicit so that you're protecting your company's risks, right?

Speaker: I know the research in terms the evidence of what that leads to, and that's good enough for me. When I'm feeling like I can control my impulses, I try to channel her because I think it's the most strategic approach. I'm not always good at it, but i'm she's my inspiration. That's fantastic.

Speaker: And I love that you talk about opportunity gap because that's very neutral in a way. Yeah. What are you reading that you would recommend? Well, I just read Super Shifts and it's written by a behavioral scientist, Dr. Janae Duane, and a futurist, Steve Fisher.

Speaker: Which, honestly, like no shade to Steve Fisher. I've never met the guy. I used to think futurists were kind of like astrologers. Like a really horrible thing to say and probably should be edited out.

Speaker: But I did. And I didn't really get Right? Right. But I received this book and it's all about how ah AI transforming how we live, learn and work in the age of artificial intelligence.

Speaker: And it provided future scenarios that I never would have imagined on my own. And so now I get it. Now it's like people who are able to see possibilities that are maybe far-fetched, but maybe it's just our own limitations in terms of imagination that makes them seem far-fetched.

Speaker: And so to be able to open myself up was a challenge because I'm kind of inherently a little bit skeptical by nature, but it did me a lot of good. Yeah. I'm now very interested.

Speaker: And here is my bonus question. What is your favorite myth about AI that you like to bust? Well, we already covered two of them.

Speaker: But I think the one to that bears repeating is that it's that it is a is somehow a universal source of truth. It's not. It's a statistical model.

Speaker: It doesn't know anything. It doesn't reason. it processes statistics. And in so doing, it's going to make a lot of mistakes.

Speaker: And on a personal level, people have to check the results of whatever they produce using ai and be skeptical. at an organizational level, people need to know that those mistakes can seriously undermine different populations based on historical inequalities.

Speaker: Very helpful reminder at the end of this episode. Do you have any final thoughts or advice for fellow changemakers or an ask of the listeners? Yes, I do have an ask because um I realize i'm start I'm leaving on a downer.

Speaker: I want every listener out there to identify a way in which they could imagine AI being helpful, generative AI being helpful for them. I really encourage you to look at Claude. It's awesome.

Speaker: Claude.ai. And then try it and then talk to someone else about it who's skeptical. You're right and tell the person you're right to be skeptical. Absolutely. Be skeptical as long as the day is long.

Speaker: And if you're skeptical, talk about it. Talk to other people about it. But don't shoot yourself in the foot by resisting it altogether because there are ways that you can be involved. And to be involved, you have to know the basics of using the tools. Yeah, absolutely.

Speaker: Absolutely. Mara, wow. I absolutely adored this conversation. It was so fascinating, so insightful. Thank you. Thank you so much for your time today.

Speaker: Thank you, Crispina. It was so good to see you. Yes, absolutely. You too.

Speaker: If you'll stay with us, thank you so much for the time that you took to discover one of the amazing women changing finance. If you enjoyed this episode, please share it with your friends, colleagues, and on social media.

Speaker: Your feedback is also incredibly valuable. Please take a moment to leave a review on your favorite podcast platform. It will help me for the next episodes. And don't miss out on future episodes.

Speaker: Subscribe to the podcast and follow me on LinkedIn. See you for the next episode.

Speaker: Women Changing Finance is part of the imp Impact Alpha podcast network. Smart conversations by and for impact investing professionals.

Speaker

Speaker

Speaker

Speaker

Speaker

Speaker

Speaker

Speaker

Speaker

Speaker

Recommended