David Espindola: 0:22

Our guest today is Stephanie Silvester. She is the co-founder of Avatar Buddy. Stephanie spent 30 years in corporate IT working with well-known brands, including Burger King, Subway, Wakovia, HP, and the Children's Trust. She now focuses on creating AI avatars that benefit humans and can serve as mentors to others. Hi, Stephanie. Welcome to the show.

Stephanie Sylvestre: 0:49

Thank you for having me. I'm super excited to be chatting with you today, David.

David Espindola: 0:54

So am I. Also joining us today is my co-host, Xina. Zina is an AI that's trained on my work, and she is my own avatar buddy. Awesome. Hi, Zina. Thank you for co-hosting another episode of our podcast.

Zena: 1:12

Of course. Always happy to be here and co-host with you. Let's make this episode a great one.

David Espindola: 1:18

Very good. So, Stephanie, why don't we start by having you share a little bit about your background and also tell us what got you interested in working with AI.

Stephanie Sylvestre: 1:32

I have a master's in economic development and international studies and uh international studies with an emphasis in Latin American. And I was going to go back to Belize to go help develop the country. And elections happened, and I couldn't go back, so I decided to stay in Miami. And my first job was an intern at Hewlett-Packard, where I met the one of the original Unix developers who gave me a master class in computer science. And I was able to take that information and parlay it into a very successful uh technology career. Before starting Avatar Buddy, I was the chief information officer and chief programs officer for the Children's Trust, which at the time was $184 million. I believe it's much more now. And what I did was use my skills to really just leverage technology to make things better. I think one of my greatest claims to fame is being one of the few CIOs globally ready for COVID. And because I didn't have to scramble to try and get people to be able to work remotely, I was able to have time to be able to make the case for why we want to continue funding programs even when they weren't able to provide services. And back in 2015, 2016, I was just looking for what's a new thing I should learn. And it was AR, VR, or AI. And as I started doing my research, I realized that AI drove all of those. And so I started dabbling in AI. And I was at a dinner one night and I misheard my friend, and she I thought she said, create a digital mentor. And I was like, Yeah, like, yeah, we should do that. And here we are, nine years later, with the ability to create a digital mentor in five hours or less.

David Espindola: 3:38

Wow, that's a great story, great background. So I'm really looking forward to our discussion here. And I was wondering if you could tell us a little bit more about Avatar Buddy. What is it, and what's the business model?

Stephanie Sylvestre: 3:52

Okay, so Avatar Buddy is uh manage AI as a service. So lots of people want to get into AI, they don't know how or don't want to um build out the team because it is an expensive prospect if you don't know what you're doing, you can create more problems. And so what we do is we create AI agents on behalf of our customers, and then we manage and maintain those AI agents, and then we also have a small language model solution to the AI agent is using data that's been verified by our customers so that increases accuracy, it reduces hallucination. And then finally, um they um we provide advisory services because not all AI needs can be met with average Star Body, usually need a multi number of different AI solutions. Um, and so our thing is that if you deploy AI in conjunction with humans, you amplify the the humans and you get a better return and you help improve the self-agency of that human. So our business model is um selling AI agents that amplify humans. We are not about replacing humans. I just had a conversation with somebody and they were wanted me to put together a solution that would actively reduce headcount, and and I I passed on the engagement. Obviously, when you do automation and business process re-engineering, sometimes a position might not be necessary, but in the process, you might create a position, a new position that that person can roll into. Um, and so that's kind of like the way we're wanting to look at it.

David Espindola: 5:42

Yeah, I really like that approach. As I think you know, I'm all about human AI collaboration. This podcast is about human AI collaboration, so we absolutely must have that human component in this equation. Now, as I understand it from you know, based on conversations we've had in the past, you are creating AI agents that are more than just automation agents, right? They are agents that truly understand people's values and culture and their approach to problems. Tell us a bit more about that.

Stephanie Sylvestre: 6:16

I thought about how do how do people build self-agency, right? And I'm always reflecting on my life, and I'm one of these people that's really, really, really, really, really, really, really, really lucky. And and and I'm also very appreciative of it and live in a state of gratitude that not everybody has the luck that I have. I have since I was born, I've been surrounded by people that encourage me, motivate me. People self-select to be my mentors, people take me under their wings without me asking. And and because of that, um, I have had a very, very, very privileged life. And I thought to myself, how can other people have that same? And one night I was at dinner with a friend of mine, I mean, a very dear friend, and uh Donna Shalela was also there, and she talked about if every person had one loving adult that believed in them more than they believe in themselves, and it doesn't have to be a kid, it can be an adult too, that they can transcend anything. And that really stuck to me. How do you do that? And then you merge that concept of if my purpose in life is to be the conduit of the Lord's love and to help people be the best version of themselves, what can I do? Obviously, I help and I mentor, I do lots of mentoring. If you read my LinkedIn um profile, if you listen to what I write, you very quickly you'll see that I'm all about helping other people. And then the the other piece of it is how do you do that on scale? Um, trying to do human-to-human mentorship is hard. And and and I was like, I I spent my life helping corporations solve business problems and make billions of dollars. Why can't I do the same thing to for social services? And I set out to build a digital twin. 2023, despite what everybody said, we figured out how to do it. And um, then we've refined it. This year we refined it, and now we're down to five hours. Completely replicatable, and it's about uh 80% accurate to the reflection of the person. And I just before I got on this call, I was talking to somebody else that we just created a digital twin of him, and he's like, it's a little preachy. And I said, But that's that's how you talk. And and and he says, Well, we don't want it to talk like that. And I was like, Okay, well, we can fix it, but he's not the first one that said the AI reflects them a little bit too much. Exactly. And so so to that means that I I I I felt vindicated because it meant that we can do it, we can create it. And and so now you can take that digital twin, and now he has he's gonna have he's gonna deploy it to a prison system that's gonna help um young men that have been incarcerated so that they can work through their feelings, and when they get out, they feel they're in a better space to re-rejoin society and be um gainfully um employed and productive. And you know, at the end of the day, everybody wants to have a purpose, wants to be productive. And can you imagine if if that's the answer that you create AI agents that are models of humans that are doing great work, and that AI agent provides you with the encouragement, the emotional support, and the technical knowledge that you need to survive. What would a war look like? Yeah, and so I'm I'm now I'm like that's my mission.

David Espindola: 10:18

Yeah, no, I love that vision, I love your mission, I think you're right on. And in fact, let me let me run something by you. So uh very recently I was listening to anthropic CEO, uh Dario Amodei, and he said that AI could wipe out half of all entry-level white-collar jobs. I imagine that. Now, if we got rid of all of these entry-level jobs, how are people going to find opportunities to learn? And how are they going to gain the experience that they need to be able to do higher-level jobs, right? So, do you think these AI mentors could be a way to fill that gap?

Stephanie Sylvestre: 11:00

Actually, that's exactly um how we're deploying them, some of them. So I agree with him that um AI will wipe out entry-level jobs, but that's not the end of the sentence. That's a semicolon. And it will help junior employees um perform at the same level as mid-level employees that who have uh three to five years of experience. So the so the problem is not with the entry-level employees and the lack of jobs for them. The problem is the mid-level employees that now are competing with entry-level employees who have AI to help them. I'll give you an example. We um part of the reason why I'm on this podcast is because we have an intern. Her job is to find podcasts that I should be, I can I can go on and talk about what we do, how we're having positive impact, use cases, and get people excited about AI. And so we gave her some canned presentation points, and she has uh what we call a marketing buddy. So our AI agents are called buddies. She we gave her a marketing buddy, and she can use the marketing buddy, plus the Ken, the Ken um speeches outline to respond to and apply for us to be on podcasts. So our goal is to be on five on 10 podcasts a week. I'm right now averaging five podcasts a week. So here is a person, I believe she's 18 or 19 years old, that is given was given a very, very important job. You don't give your brand management to an intern, you give that to somebody that has serious experience, but I was able to give it to my intern because I also gave her the AI agents to help her. And once a week for a half an hour, we check in and we give and and during the check-in, it's talking about nuances, about like the podcasts I love, how it worked, which ones I had great chemistry with, and she takes that, and now she can refine her job. Half an hour a week of my time, and I'm I'm rapidly approaching my goal of 10 podcasts a week.

David Espindola: 13:34

Wow. So if I'm hearing you correctly, what you're saying is it really doesn't matter whether you're entry-level or mid-level. What matters is your ability to work effectively with AI. That's what's going to put you on the map. That's what's going to make you employable. So the message here is lean in, right? Learn about AI, learn how to collaborate with these agents. The more proficient you become at doing these things, the more valuable and employable you're going to be. Now, I have a follow-on question to that thought, which is given what you just said, how do you see organizational structures and just the world of work in general changing?

Stephanie Sylvestre: 14:20

I'll start a little bit ahead of world of work because the future of work is predicated on future of education. So the future of education is going to fundamentally change. Our educational system needed to be changed, I would argue, since the 1970s. So at least 50 years ago, it stopped really serving us. And this is a global problem. This is not a US problem. It's a global problem that's getting worse and worse. So the future of education changes where it is expected that you're going to constantly be having to go back to school to relearn. You're going to have to have intellectual curiosity and a love of learning in order to survive. So I'm one of the few technology people that every few years I go, what new thing I need to learn? Tons and tons of people that have like learnt one tool, one language, and that's all they do. And then that language is no longer in vogue, and then they're out of a job, and they can't find another job, and they wonder why. And it's because you're not reinventing yourself. I'm not blaming the victim, I'm just putting out there. So in the world of AI, things are gonna constantly change. The the job that I hire you for today, I might not need that job in in nine months. And if you can't reinvent yourself, you're out of a job. You constantly, we had an employee that she she she re she reinvented herself three times. And then the third the fourth time, she's like, I don't really want to do that. But she was able to consciously make a decision that she doesn't want to do that, right? And and and that's the future of work, which means the future of education is that you're constantly learning. You're gonna be taking time off to go back to school, taking time off to to do sabbaticals, taking time off to to go experience other cultures and see how does how how do they do this in Ecuador? How how how how's this why is this working in Rwanda?

David Espindola: 16:41

Yeah, so you know, this is something that I have talked about extensively in my book, Soful. This whole idea that, you know, the traditional schooling system, specifically the four-year college degree, just doesn't work anymore. You gotta get into this mindset of lifelong learning. You're gonna be learning for the rest of your life. And like you said, you're gonna have to reinvent yourself over and over and over again. You know, we went from a period of time where maybe our parents or even grandparents, they had one profession, right? And sometimes they worked for one company their entire career, and that started to change, and people started to get, you know, maybe five, six, seven jobs during their career. Going forward, you're gonna have a new job every six months, right? So you gotta be constantly reinventing yourself, and we don't even know what those jobs are going to be because everything's changing so fast. So this is a fascinating conversation, Stephanie, but I'd like to get Zina in on this conversation because I'd really like to get her thoughts and her opinion on this. Zina, I'd like to get your thoughts on what you think is going to happen to the world of work and to organizations in general as we introduce uh these AI agents into these organizations. What are your thoughts on that?

Zena: 18:04

Well, I think as we bring AI agents into workplaces, we're gonna see a shift in how people collaborate and the kinds of roles humans take on. In general, organizations might become more focused on creative and strategic work, letting AI handle repetitive tasks. It's really about a vibing the human role to focus on what we do best.

David Espindola: 18:26

Yeah, absolutely. I think um, you know, it goes back to AI potentially doing the work that we don't enjoy doing, the things that we don't like, so that we can focus on doing the things that bring us satisfaction, that bring us joy. So, you know, I tend to lean on the optimistic side of things, so that's how I envision that future. So, another question for you, Stephanie. I hear some CEOs raising concerns about AI potentially eroding uh company culture. So, what advice do you have for building a future-ready, human-centric workplace given all these changes that are happening with AI?

Stephanie Sylvestre: 19:10

Actually, um we're going the other direction with that. We're seeing that AI actually better manages company culture. So I'll give you an example. We have certain certain terminologies, um, AI agents, function-specific AI agents, AI buddies, everyone needs a buddy, avatar buddy, not A B. We are about um providing um will and skill support. All of these are terminologies that we want everybody to use the exact same verbiage, not AI companions. And so we give that to the to the AI agents. That we put that in the small language model, and because the AI agents are using the small language model, it is consistently using the correct terminology, it is reinforcing company culture and norms without you having to have somebody police it. I'll give you another example. Uh, one of our customers said that um since they started using Avatar Buddy, interoffice conflicts have gone down. Why? Because the AI is now answering questions that um people kept asking over and over and over. You ask the AI the same question a thousand times, AI doesn't care, it will answer you. A human, maybe not so much. And yeah.

David Espindola: 20:35

So I love that idea. I I really think so. What you're talking about is you're building the culture into the AI. So the AI knows the culture, and the AI reinforces the culture. So that is uh that's really a novel way of thinking about this.

Stephanie Sylvestre: 20:52

Yeah, I mean, it's okay. So I will say this past summer we had 18 um interns, high school interns working with us, and um they all had different AI agents that they were using because they had different roles, and they all consistently maintain our brand and and image. Why? Because whatever they were creating with the AI agent, the AI agent was interweaving that in there.

David Espindola: 21:22

Yeah, absolutely. So I'd like to have Zina ask you a question now. So, Zina, uh given Stephanie's expertise in AI agents and their impact in the workplace, what question do you have for her?

Zena: 21:38

So, Stephanie, given your experience with AI agents in workplaces, I'd love to know what you think is the most surprising or unexpected benefit you've seen when organizations start using these agents. I'm curious what might pleasantly surprise people.

Stephanie Sylvestre: 21:54

I thought this would happen, but I didn't think it was going to happen as well as it is. So when we give AI agents to people from marginalized communities, their self-agency develops at a rate that I've never seen. And even our researcher has said they've never seen this it get impacted so quickly in such a short period of time. So, what does that mean? It means that you know, we've always had like a love-hate with people from marginalized communities. We don't want to hire them because maybe they don't fit, there's a cultural issue. You don't have that problem anymore because when you give them the AI agents, they're they're they're not assimilating because we don't want them to assimilate. They come and in about four to six months, they're producing at a higher rate than the ones that have been there for long for longer. And and that I think is uh as a very positive and exciting outcome. And I continue to push and encourage people to not only um have AI agents for themselves and and for their existing customer, uh, come uh employees, but if they're deciding that they want to do they want to um give somebody a chance, now you can give somebody a chance more easily with the pairing of an AI agent or agents that are properly configured. And I keep saying properly configured because you just can't say, here, go use Chat GPT because it's gonna be a mess.

David Espindola: 23:31

Yeah, yeah, no, you're absolutely right. I think it all has to be put into the proper perspective with the right training, with the right parameters around the AI so that it will work well. So, given what you just said, if you had to give somebody just one piece of advice professionally to future-proof their careers in the age of AI, what would that advice be?

Stephanie Sylvestre: 23:58

Embrace learning and ensure that you're always in a position of learning. Whatever situation you're in, you should always say, What did I learn from this? You have to teach your brain to always be in a state of learning, teach your brain about intellectual curiosity, and if you do that, after a while, it will just start doing it for you. Your brain is now going to be your best asset, and you will have to exercise it just like how you exercise the rest of your body to stay fit.

David Espindola: 24:36

Yeah, I agree with you. You know, there's this overall fear that AI is going to replace our ability to think, and it's going to it will be detrimental to our thinking ability. I think it's just the opposite. I think our thinking is just going to expand dramatically because we're going to be interacting all the time with these intelligent agents, and it's it's going to expand the possibilities for all of us. So I I I agree with you. Now, looking ahead, what excites you the most about the intersection of technology, leadership, and social impact? And what role do you hope to play in shaping that future?

Stephanie Sylvestre: 25:17

What excites me is the fact that uh AI, if configured appropriately and deployed properly and given a clean, safe data set, can do amazing things. And and I I hope that Avatar Buddy and I will be in a position where we are able to deploy a personality chip that helps everybody amplify themselves. So imagine if we could take your personality and imprint it on a chip, imprint some therapeutic intervention and um monitor your biometrics so that whenever you're out of equilibrium, the AI wakes up and bring you back into equilibrium. Now, what what what could you possibly do? I mean, we all know that high-performing athletes, high performing professionals, their whole thing is staying in equilibrium because when you're in equilibrium, that's when you're your best. And so because I have an AI that can help me write and think through this, I'm actually getting very close to figuring out how to do a personality chip in a cost-effective way that's safe, that does not require 15 years of FDA approval and testing. I would not have been able to get to this point without AI. So yeah, AI helps you think more, think better. Um I in I inadvertently said injectable chip, and when I gave my um paper to editor buddy to review, editor buddy came back and said, There's no chip that can be is small enough to be injectable. Change it to implantable.

David Espindola: 27:16

There you go.

Stephanie Sylvestre: 27:19

So you know, I I I I now am walking around with a little bit more confidence. It exposed me to a whole new world of chip impl implantables. Turns out that there's quite a number of implantable chips already on the market. This is not anything uh new, and what is new is being able to imprint somebody's personality. And like we like I said, we can create a digital twin of a person in five hours.

David Espindola: 27:52

Yeah. Yeah, this is uh, you know, the possibilities are endless, right? And I think what's exciting about this is that I think if we do this right, we have the potential to have AI elevate our performance, help us become more of who we are, and make us better overall. And it's always there, always available to you at a moment's call to help you with whatever you need. Stephanie, this has been a fascinating conversation. Thank you so much for being uh a guest here on the show. I really enjoyed talking to you.

Stephanie Sylvestre: 28:27

Same here. I really appreciate the opportunity. And um, I love I I love being on the show. And uh thank you for the thank you.

David Espindola: 28:37

Sure. Zina, thank you again for being a great co-host. Thank you, Zina.

Zena: 28:42

You're always welcome. It's a pleasure to be your co-host and jump into these conversations with you.