#370 Redefining Learning: Juliette Denny on AI, Gamification, and the Future of Education

Mehmet: [00:00:00] Hello and welcome back to a new episode of the CTO show with Mehmet. Today I'm very pleased joining me Juliette Denny. Juliette, the way I love to do it is I keep it to my guests to introduce themselves because I know no one else can talk really about someone else better than themselves. So the floor is yours.

Juliette: Fantastic. Well, Mehmet, thanks very much for having me on. Very much appreciate it. Um, my name is Juliette Denny. Um, and at the moment I wear, uh, three different hats. So I am still involved in a company called growth engineering, um, which, um, has been around now for 20 years and we focus very much on, um, helping large corporates deliver learning solutions to their, um, to their teams worldwide.

Juliette: So, uh, our clients are people like, uh, Apple, L'Oreal, these, these kind of folks who are looking to do something kind of gamified and, and, and social with regards to how they deliver out learning. So that company is a kind of learning management system. We also have learning [00:01:00] apps. Um, and, uh, we also have, um, authoring tools as well.

Juliette: Um, so that company's been 20 years. I can't believe it. I cannot believe, uh, you know, how quickly that time has flown by, but yeah, 20 years, uh, we were in March. Um, and still, you know, kicking ass and, and trying to, um, you know, change the way corporate learning works. Um, but my new, um, is a company called Iridescent Technologies.

Juliette: Um, and we are in the process of launching a new kind of learning companion called Zavmo AI. Um, so it's due to come out at the end of, um, this, this, sorry, next month. And what we're really trying to do with learning is really kind of break down the, um, going from one to many. So typically what would happen, a teacher would stand up in class and basically, you know, would deliver a lesson to a one to one.

Juliette: So this applies for e learning as well. So you don't build it once and then [00:02:00] basically it gets delivered out to everybody. The idea was Zavmo. is it's individual conversations, and I'll go into that a little bit more. But it's just been completely fascinating. And then the final thing that I'm doing is it's bringing out a book called, um, the learning revolution, which is again coming out in a couple of months.

Juliette: Um, and so very exciting and a brand new challenge for me. I've never written a book before, um, but just talking about how learning will be affected by, you know, the coming new wave. So, yeah, that's me in a nutshell.

Mehmet: Fantastic. And again, thank you, Juliettete, for being with me here today. Now something You know, in some situations I'm curious about because I spent a little bit of my time during my career in that space in academia, but for you, it was more into an L.

Mehmet: N. D. And when I shifted, you know, companies across corporations, you know, it was a major department, you know, when it comes to building I mean, course materials, any [00:03:00] learning material for people who work inside the organization. And to be frank with you, it's not that easy task. And this is very complicated.

Mehmet: So out of curiosity, Juliette, what brought you and attracted you to be in that domain?

Juliette: Yeah. So learning and learning and development, um, it was a bit by accident to be honest with you. So this is obviously a long time ago. So cycle back, you know, uh, You know, 20 years ago. Um, my I started my career in sales.

Juliette: Um, and my first role was in, you know, selling box software. Um, I then my husband and I moved because we were doing our post grads at Harvard. So we then moved to the Boston area and I joined dot com startup again in a kind of business development type of role. Um, and I, I, I really enjoy, you know, business development, but I knew that when we came back to the UK, I wanted to do, I wanted to basically have a, um, uh, a technology company.

Juliette: I was absolutely transfixed by this idea that, and I guess it [00:04:00] was because I was in box software, which obviously if you make, um, you know, if, if there's a bug or something like that, then you have the, you know, put a patch on it and it's a complete nightmare. It was completely trying to fix by this idea that you had a set of code.

Juliette: It was hosted on the internet and you could basically deploy on demand. I was just like, that is just insane. Build it once, sell it, you know? Um, so I came back to the UK, very, very clear that I wanted to build a SAS software company. Um, And in order to sort of kind of get into the market, um, and because I've been involved in the, the.

Juliette: com boom and bust, um, I started, uh, doing a lot of consultancy with banks and VCs, looking at their technology portfolio, and then basically making recommendations to them. to as to whether they were likely to be able to hit their targets or not for for further round of funding. So it was a kind of role where it was like, is the technology ready for commercialization and or if it was, if it was ready for commercialization, what did they need to do to move from [00:05:00] kind of a tech founder mentality and a sort of a technical focus, product building focus into, you know, the commercialization phase.

Juliette: And so through that there were Lots of problems that were identified in terms of, uh, you know, understanding commercial process, uh, but then also the skills base you need to go out and sell your product. Um, and so what I started doing there was I got a developer and, you know, we started coding up, um, you know, ways to automate this.

Juliette: So if I had, you know, looking at their, their, their processes and then delivering content on the back of that, and then from that morphed. the growth engineering, you know, the growth engineering, uh, you know, platform that, that you kind of see today. Now, I think what, what was really interesting is at the time, because we were delivering out commercial people, we really needed to make the learning experience as fun as possible.

Juliette: So very quickly, you know, we had things like gamification, um, and social. And so [00:06:00] because of that, we were able to, You know, attract some really big clients quite quickly to the company and then, you know, grow as fast as we did. And so far, I think we've won about 160 global awards. You know, we've worked with and we still do work with, you know, some of the biggest companies in the world because I think what we do is just quite unique.

Juliette: It's very, very learner focused. Um, when the learning management market first came into, uh, evolution, it was very much about reporting like what people did. And so the quality of the learning wasn't really that important. And so what we really tried to do is really kind of flip that on the head and say, look, learning is all about the learner.

Juliette: It's only, you know, only going to get behavior change. If the knowledge that you're delivering to the learner sticks, so how are we going to do this? And so, you know, our, um, our mission is very much to, to use, you know, brain science and neuroscience in order to, you know, lift the quality of the learning [00:07:00] experience.

Mehmet: Absolutely, you know, fantastic journey, I would say also as well. Now, one term you used and, you know, I'm familiar with it, but I'm going to ask a little bit to maybe explain it very shortly, which is gamification in learning, right? So first, what is it like in a simple term? Second, why is it important, you know, to, to incorporate, you know, gamification in, in, in, in a company learning?

Mehmet: And maybe you have seen it, you know, from both sides, Juliettete. So because you just mentioned, you know, you work with big names like Apple, L'Oreal and others. And at the same time, you know, I'm curious to know if like such techniques like gamification can fit even like smaller companies, maybe even startups.

Juliette: So, so first question, what is gamification? So gamification is different to game based learning. [00:08:00] So game based learning is turning learning into a game. Gamification is using gaming mechanics. Things like streaks, points, XP. levels, leaderboards, uh, peer to peer battles, those are gaming mechanics to make the learning experience more interesting and more exciting.

Juliette: So, um, it's quite difficult for a company to turn every single piece of learning into a game because it's by definition incredibly, it can be quite expensive. Uh, not now with AI, not now with my new company, but you know, in the past it was. E learning can be very, very linear. Um, and there's a beginning in the middle and end and you have to do, you have to outline the learning objectives.

Juliette: Uh, then you need to, uh, you know, Meet the learning deliverables. The learning need to meet the learning learning deliverables, and that can be quite complicated to put in the game. Not [00:09:00] impossible. So, you know, I'm currently working on a project with Moody's, which is which is very, very, very like that. So it's not impossible.

Juliette: It's just very high performance, very high, you know, high production values. So gamification is basically using these gaming mechanics which can be embedded onto a platform so that any piece of learning that you put onto the platform is by nature gamified. You are giving out XP, streaks, whatever it might be in order for when the learner does something.

Juliette: Um, and with gamification be used by smaller companies? Of course. Because we see Duolingo or, you know, even Zimu, you know, the, the, um, Chinese shopping app. So, you know, gamification is just a mechanic, um, and it's used by individuals all the time. Um, so yes, it's, it's, it's, it's, it's, it's, it's the reality, yeah.

Mehmet: Cool. That's good to know Juliette. Now, one thing every [00:10:00] time you mentioned, I see, you know, like you are getting excited. And of course you're building something around it. Um, you know, transitioning to AI powered learning, right? So tell me a little bit, walk me through when exactly you felt, okay, now it's time to take this track.

Mehmet: And what makes you excited about, you know, embedding AI with learning and development?

Juliette: So I've really, I really love this question because it, it stems from, um, really me being a little girl and, um, my learning experience. So the reason why I am so excited so fascinated and interesting, interested in learning is because I'm dyslexic.

Juliette: Um, my kids are dyslexic and ADHD and I have always felt that [00:11:00] I did very, very badly at school. So, you know, I was, Um, you know, didn't, didn't, didn't make it, you know, past my sort of high, didn't make it, make it through high school eventually, you know, very, very slowly. I managed to sort of make it up and ended up, you know, uh, you know, studying incredibly hard, you know, Harvard night school.

Juliette: And so I was able to get, you know, my qualifications through Harvard. But, um, I think for a lot of people, they are locked out of education. And I feel like that is a burning injustice in my soul. So I feel that everybody on this planet has the capacity to achieve extraordinary things. If they are taught in a way that works for them.

Juliette: So if you are dyslexic, you have, by definition, short term memory. I have short term memory. Which means that, you know, learning by rote does just not work for me. Right? It just doesn't work for me. And people are born [00:12:00] curious. right? Every human has the capacity to learn and flourish. Curiosity is, you know, ongoing burning.

Juliette: Curiosity is one of those things that basically keeps us from aging, you know? And so my passion is not necessarily for technology or for AI. My tech, not my passion is for how do we get technology and now AI to deliver the learning experience so vastly better so that every individual can be inspired to basically deliver their best life.

Juliette: you know, that to design the life that they want, and that's really where iridescent technology, you know, comes in. So when you ask me what, um, you know, what, why did you suddenly change from growth engineering to iridescent technology? Very simply because I wanted a company that was born out of AI. So looking at the power of what AI can, you know, achieve in terms of large language models and knowing what I know about brain science and how we [00:13:00] curate.

Juliette: Curiosity and excitement for learning and how we get people to engage in investing in themselves as a habit. Um, I was like, AI can deliver that. So it wasn't that I was drawn to AI because of some technical curiosity. I was drawn to AI because that my burning passion is to leave the world with a better, you know, education system than it currently does Have.

Juliette: does have now. Now, what I am not saying is, is that, you know, those individual teachers that inspire and basically, and I had those individual teachers in my life that basically said, no, Juliette, you can do it. Let's just with a bit of patience, with a bit of dah, dah, dah, you can get there. But what I'm saying is by using AI, we can have a personal learning companion that can help us from the get go.

Juliette: Um, and, and using all those powerful techniques that teachers do already have and know. and embed that into the large language models to deliver something, you know, really [00:14:00] extraordinary in terms of a learning experience.

Mehmet: I love this Juliettete, because, uh, you know, this is a discussion that always, even I have it with my own daughter, like about, you know, she keeps asking me why, you know, they assume that everyone in the class, you know, they will react to a lesson the same way or everyone in the class will understand certain material and like the same on the same level and even myself, of course, like I was always feeling okay.

Mehmet: Some people, they might be good, for example, at math. And some people are not. But you know, at the same time, we get all the same materials in the school. To your point, like not all the teachers have the passions, then not all of them. They have the will to go and, you know, go the extra mile. I can't blame them, of course, because, you know, at my time, you know, we were like 25 plus people in the class.

Mehmet: So you can imagine we're like a little bit crowded in the place where I was born and raised. So I'm loving this [00:15:00] because I can see Juliette that AI can, to your point, customize the learning experience, right? So let's say if I, if I get it right from you, so the AI assess my capabilities. assess the way I can understand and process, you know, new things.

Mehmet: And based on that, it will build for me maybe a learning path. Is this, is this exactly what you're talking about, Juliettete? Like feel free a little bit too. I want you to explain it a little bit more. If there's something you, if there's something patented or something, okay. I can understand, but you know, I love to hear more about the details.

Juliette: Fantastic. Well, it's my favorite subject, so no problem at all. So we have, um, what we call a four step process. Okay. We call it the 4D process. Okay. So the first one is, uh, discover. Okay. So one of the things in order to provide a [00:16:00] personal, um, engaging learning conversation, we need to understand you. Okay.

Juliette: So the first thing that we're going to do is we're going to discover who you are. So in a corporate setting or a professional setting, this might be, tell me a little bit about your role. Tell me about how long you've been in that role. Tell me about, um, the company that you work for. Tell me about the values that you have.

Juliette: Tell me about the, um, challenges that you have in work. Tell me about the strengths and weaknesses. So in, in In in the real world, the way we would do that is training needs analysis. You would sit down with somebody from HR and they would go through and they would talk to you in exactly this way. And together you would discover your ideal career path or your ideal learning pathway to get you to your career goals.

Juliette: The next thing, so while Zavmo is, while you're going through this, this discovery process, and Zavmo's asking you lots of questions, [00:17:00] Zavmo's also trying to understand, um, the next bit, which is discuss, okay? So, the next bit, when you are presented with a curricula of learning, is habit forming, okay? So, in the discuss phase, what Zavmo is doing is Zavmo is saying, right, and You've just said, Juliette, that you want to move from, um, you know, a SAS, a SAS company and you want to start an AI company, um, and you want to learn, uh, about AI, um, and you want to learn about AI to the level that you might get at, uh, let's say MIT.

Juliette: So degree level, level, level seven. Um, You also, um, have also spoken to me in the past that you have challenges in terms of managing large numbers of people. Um, and so we're going to, we're going to throw in a module on that as well. Um, and then you've said that you would like to complete all of this learning within, you know, a year's time.

Juliette: So that will, and you've also said that because you're dyslexic and you don't have really bad short term memory and you've got, [00:18:00] you're slightly on the ADHD spectrum as well. So you don't like large chunks of learning. You also want to learn 10 minutes. It's in no longer than 10 minutes in a 10 minute burst.

Juliette: So that will mean, Julia, that you need to, uh, check in with, with, with Zavmo, with me as your learning companion, at least every sort of 10 minutes. So that discuss phase is about agreeing what your habits are going to be like ongoing in order for you to complete what you said that you want to do. And that's really important because in learning at the moment, you don't really have the choice.

Juliette: Okay. Thanks. of learning what you have to learn. It's told to you that you have to learn it. Even in a corporate setting, they'll go, okay, well, we're doing this sales and marketing management level four certificate, and you've got to do that. Or you've got to do health and safety and you've got to do it and you've got to do it by this date.

Juliette: Okay. So what that does to the brain is basically the brain goes, this is something I have to do, not something I want to do. So that disgust phase is really important in terms [00:19:00] of how you engage and how delighted and excited you are. to actually do the learning. Okay, so that's stage two. The second, the third stage is deliver.

Juliette: So we're now going to go in and we're going to, Zapmo is going to start delivering your learning experience. Now, from, from the get go, okay, we solidly have the, um, certainly an MVP. We have the, um, the, the commitment that learning is a conversation. Okay. So learning should not be I talk, you receive it just in the same way as we're doing now.

Juliette: I'd like to think that you're engaged in this conversation because you're asking questions. I'm also asking questions and together we're able to move towards an understanding of, you know, you were explaining a little bit about how the CTO show happened and I'm understanding that. And it's, it's a joy for me because I'm curious.

Juliette: I'm in receptor mode. I'm also in, you know, delivery mode. And so that [00:20:00] interaction is very important. So Zafmo is not. A, here's a block of text. Here's a video. Here's a, it is a conversation. So Zavmo might say, Hey, look, you know, check out this two minute video and then we're going to have a conversation with it.

Juliette: But in the main, it is all conversation. So it is basically what Zavmo uses. the company that you work for, the, you know, the industry that you work in, the role that you're in, uh, the subject area that you want to study. And it combines all of that into a healthy conversation. And so basically what you will either be talking into Zafmo or you will be typing into Zafmo further down the line, we'll be able to do it via video.

Juliette: But this discussion is really important. Now, I briefly mentioned the idea about how learning qualifications work. So. In the UK, in India, in the US, across Europe, you have something called, um, uh, you know, official qualifications, um, frameworks. And so what ZAFMO does is it pushing out to these [00:21:00] official, uh, qualification, uh, networks to make sure that the overall quality of the learning conversation meets the standards mandated by those national governments.

Juliette: So it's looking in the, um, delivery stage at, uh, It's looking at learning, um, objectives. So what are the learning objectives? So what do, what does the curriculum need to teach you, okay, in order for you to understand that, that module, that certificate, that award, whatever it might be. And then finally, We go to the fourth stage and the fourth stage is called demonstration.

Juliette: Now, demonstration is really important because when you are learning, you are not really that interested, or most people aren't that interested in the receptors. They're receiving information where learning becomes interesting for the brain, where our synapses start kind of, you [00:22:00] know, is in the application.

Juliette: So if you look at learning taxonomy, things like Bloom's taxonomy is really, a really good model to understand this. What Bloom's taxonomy shows is we become more committed and more interested in the learning that we're doing, the higher we move up Bloom's taxonomy. So the more we start applying, analyzing, rather than repetition, you know, reinforcement.

Juliette: We're not really that interested in the lower areas of, of, of, of course we have to do them in order to get to the higher areas of taxonomy. So what Zavmo does in the final stage is Zavmo gets you to demonstrate Now, how Zavmo does that is through case studies, real world scenarios that again, mapped on to everything that Zavmo has come to understand about you throughout the 4D process.

Juliette: The other thing that's really important about how Zavmo asked that [00:23:00] question is that is that it's based on the learning outcomes that are mandated in the qualification that you're studying. So what that allows ZavMo and Iridescent Technology to do is wrap all of that up and say, here, National Awarding Body, this person has actually successfully achieved the outcome.

Juliette: They've just, achieved it in a fundamentally different way. But in terms of, you know, learning credits, learning hours, all of those kinds of things that go into recognizing whether you have actually achieved the full spectrum of qualification, it, it can show that it can show that you have. So, you know, we're tremendously excited about You know what, what this could do.

Juliette: Um, you know, there are loads of other things that it does. So it also takes in things like narrative. So one of the ways that humans learn a lot better is through the power of storytelling. And so the more Zafmo understands about you, as you move [00:24:00] through learning, not only can it weave in, you know, You know, examples from your, you know, the company that you work in and all of that, it can also weave in narrative and story in and around things that you are interested in.

Juliette: So fundamentally making the, the, the learning more, making you more curious about your learning.

Mehmet: That's very interesting. Really, really interesting. And I can see it being applied in different areas, like not only for, I mean, corporate learning, so it can even help someone. You know, you know, we we see sometimes on the internet like these people who are looking for a change in their career because they discovered after working for a long time, let's say, I'm just giving an example in, in accounting that, you know what?

Mehmet: Like, this is not actually my passion. My passion is X, let's say, and let's say this X is maybe, I don't know, tech. And then they say, Oh, we [00:25:00] would like to build things. We would love to do that. And then, you know, like we see they put a lot of efforts to go and learn, for example, coding and so on. But maybe with what you are describing, Juliette, it's also give them the credibility of doing this when, for example, changing careers or maybe if they want to move.

Mehmet: up in the ladder and, you know, maybe acquire something related to leadership. Correct me if I'm wrong. Like, I'm just, you know, trying to, you know, what I like to do usually is when someone is explaining a very exciting thing, where I can, where I can put it, like where I can see like immediate use of it. So it looks like something really, really interesting, Juliettete.

Mehmet: Um, Have you talked to anyone about it? Like how people are getting, you know, excited about something like this?

Juliette: I mean, I think people are in the main they are very excited about it Um, but of course it is all very new. Um, and so, you know, although i've i've seen it on the back end Um, we are not we're not going to launch beta until the end [00:26:00] of july So we've been working on it and it will launch at the end of july um I think um The the idea For most people is, you know, is a utopia for education because basically what it means is, is that they, they will be able to fully unleash their, their potential.

Juliette: So I think, you know, there's no doubt that it's going to take us, you know, a lot of iterating. And so, you know, having, you know, grown growth engineering over the last 20 years, I have no doubt that we will, you know, come out of the gate with, you know, a perfect product. I would love to think we have, and certainly on the back end is looking, you know, it's, what's incredible is, is that it, the way it makes you feel, because when you're talking to Zavmo, um, It kind of goes at a pace which makes you feel really confident.

Juliette: So you want to move forward more and that, and that for me, I think is really interesting. So when I'm using [00:27:00] other AIs, so I use Gemini a lot. I use, I don't use chat GTP, but I use Gemini a lot. Um, what, what I find is, is that it's kind of like a glorified search engine. Do you know what I mean? It really is.

Juliette: It's like you ask it a question. It does it, it might do it in longer form, but it's not, it's not right. And I understand automation and I'm not in that business. Yeah. So I understand that AI does a lot more than just. Do what I want it to do. But I think when you see it in the context of what we are trying to do it in, that's really what makes me think, wow, what an incredible technology.

Juliette: Um, because, you know, it, it, I think it's the flexibility of it as well, so that you can, um, you know, in ensure that the guardrails around the, um, the, the, the learning conversation are. are framed in a way that we already have within the learning and development environment. So because we can use a kind of deep understanding of some of these, um, learning pedagogy models, like [00:28:00] Bloom's taxonomy, like, um, you know, the forgetting curve, like, you know, um, you know, open loop theory, like, you know, there's lots of these theories that we can basically get ZAFMO to understand, Now, because Zavmo understands that, it kind of puts guardrails around the AI to ensure that the experience is one that you continually want to go back to.

Juliette: So, it's so easy, it's so interesting to see how you can mold AI. And people all, you know, talk a lot about how scared they are of AI and how it's going to take over the world and all this kind of stuff. But what they don't talk about is the flip side of that. The fact that AI is here to facilitate the absolute best.

Juliette: that humans can be. And, and, and that's, I think what the cusp that we're trying to do. We're trying to take this incredibly interesting technology and frame it in a way that really does allow humans to really design their best life. That's really what we're after. We want every single human in the [00:29:00] world to be, you know, designing the best life they possibly can through, through Zavmo.

Juliette: At an incredibly low cost price. I mean, that's the other thing. Once it's built, it's, it's not as expensive as putting a teacher in a classroom. It's not as expensive as, you know, uh, signing up for a course online. You know, it should be. Um, uh, uh, allow, you know, for social mobility and, you know, all of the things that we want in life, which is, you know, a fairer, you know, better society, you know,

Mehmet: but yeah,

Juliette: fascinating,

Mehmet: you know, you took one question from me.

Mehmet: You mentioned something, but again, I managed to get, you know, the question rebuilt in my mind. So. Which is, I think it kind of makes sense also for, for, for the audience. You just said something like you don't use chat GPT. Don't tell me why not yet. You mentioned, and this was my question about, you know, people who are skeptical about AI, right now, [00:30:00] why Juliette we need, and I'm happy you're doing this because, you know, when we say the word they are nowadays, the first thing that comes to people's mind is just the rapper on top of The large language model that, for example, OpenAI through ChatGPT they did, or, you know, Google with Gemini, uh, Anthropic with Cloud, right?

Mehmet: So, I feel there is a need, and correct me again if I'm wrong, and I'm happy you're doing this, actually, with, uh, uh, with, with your current, uh, project, is Letting people understand that AI is not just chat GPT or Gemini or whatever, like there's a huge concept of large language model and this large language model you can train it the way you want based on some general knowledge, but I would love to hear that from you.

Mehmet: First of all, why you don't use chat GPT and you know, how we can push more to change this concept. Uh, I mean, this, this skeptical, [00:31:00] uh, look to, to AI from, from the mass audience.

Juliette: I mean, just absolutely brilliant questions and one that, you know, I think about a lot. Um, so, so let, first of all, let me ask the first question.

Juliette: Why don't I use chat? You could be very simply because my company has Google Cloud and Gemini comes free with Google Cloud. So I does what I want it to do. We do not use Gemini, I should say, uh, for building Zavmo. Okay. And, and just a slight segue, which I would love to share with you, because I think listeners will be very interested.

Juliette: Just quickly, how did I get so, how, how did this gem of an idea get, you know, move from an interest in, in learning to starting off a new company? My son is currently at. at Tokyo University. And he started getting really into AI. And we also talk a lot about entrepreneurship is our, uh, in our [00:32:00] family generally.

Juliette: And I was saying, you know, I just can't help thinking with what large language models can do. You know, we could really create something incredible in AI. And before I knew it, Cosmo, uh, my eldest son had decided that, you know, he wanted to give up university and he wanted to study engineering. Start this company with me.

Juliette: So the AI powerhouse behind iridescent technology is actually my son. Um, so yeah, Cosmos passion for, for AI is, um, well, it's just all consuming for him. So when these large language models, you know, came out, um, they, they're obviously appearing at a rapid rate and some of them are better than others. four different things.

Juliette: So, for example, Gemini is particularly good because it has a, you know, a large, larger number of index pages because it uses Google. So that's why I prefer, you know, Gemini. But these things move so quickly and so fast. So how did we end up sort of, um, I guess, coming together this idea of, um, you know, A. I.

Juliette: [00:33:00] And, uh, you know, learning pedagogy. It was really through a conversation on the four D's. So we were talking about how potentially, you know, AI could deliver this idea of, you know, discovery, discussion, delivery, and, you know, demonstration, which is really important for any kind of, um, you know, learning project that you're, you're, you're doing.

Juliette: Um, And how do we, how we would deliver that in, in, in AI. So we, so Cosmo immediately sort of, you know, got on to, um, I think we started off on Claude, uh, we started off with Claude and produced this, you know, the, the, the four D's. Um, and we were like, yeah, okay, well that's good. But there are also other elements in there that we're going to need to, to put in.

Juliette: So for example, going back to Bloom's taxonomy and going back to the areas of interest that make. So that's when we started. putting in, you know, uh, prompts in and around Bloom's taxonomy. And then we were like, well, you know, uh, we don't, we want to offer pretty much every qualification that anyone can do, uh, [00:34:00] that is recognized by their national governments.

Juliette: So then we started saying, right, okay, well, how do we feed in, via APIs, um, information coming in from things like Ofqual and EFQL, which is the European Qualifications Body. So feeding that into the model. Um, and, and, and, and so it goes on and so it goes on. So you're just constantly taking this model and then, and then what will happen rather delightfully, um, is that you'll get another upgraded, um, LLM and you'll be like, you know what?

Juliette: I think. We could actually go faster and cheaper with a different LLM model. So you'll take all of those connections, all of those prompts, and you'll basically put it into another model and it will work faster, better, cheaper. So the architecture of the way that we are creating. Um, creating Zavmo is really interesting as well.

Juliette: So obviously you've got the kind of DevOps layer, which makes it goes fast. So concurrently, we know that we can get, you know, at least half a million, uh, people concurrently [00:35:00] using Zavmo when we launch, uh, not that, you know, at launch, we'll need that many, but, you know, in terms of the infrastructure, and we've got the DevOps, we've got the DevOps layer, then you've got the, if you like the, the, um, you know, the, the, the, the kind of core infrastructure.

Juliette: engineering level, then you've got the prompt engineering level, and then you've got the large language model. So what we're hoping to do, um, once we've launched, uh, ZAFMO is move into our own large language model, which specifically focuses on, um, uh, understanding how people learn. So that's really kind of our goal, but you know, we are, you know, somewhere away from that.

Mehmet: We

Juliette: need a quick half billion in order to do that. But let's get, let's get the MVP out there. Let's get people learn, you know, using it and then. Yeah. And then, and then, and then go from there.

Mehmet: Uh, I don't do much predictions on the show much, but you know, from what you are describing Juliette, I think in no time you're going to have a good [00:36:00] traction of people using, uh, Zavmo because the way, and you know, by nature, I love everything, structured, planned and clear to me.

Mehmet: So I know how I'm going to go from point A to point B, right. And with the four D's you just, you, you, you explain, I can see something which is logical. Easy to do at the same time. Of course, there is some hard work to do sometimes. We have to do sometimes hard tasks. But, as someone, I don't claim myself having huge experience in education or anything by any mean, but the common sense in me says, like, this is something really going to become big.

Mehmet: You know, I love also one thing, Juliettete, and this is what I want to ask you. This spirit of entrepreneurship, and you just said, like, you're doing it with your son also. So how exciting is it the time now for entrepreneurs to be in the age of AI where really they can do things? And I need, you know, kind of advice from [00:37:00] someone who've been in the tech space for more than 20 years, Juliettete.

Mehmet: You know, I want, Your point of view on what are the limits that entrepreneurs should be aiming for more than just putting, you know, a wrapper on top of a API and just making like chat bots or whatever. I want to hear your feedback on, on, on that part.

Juliette: So I think this is a really interesting debate and I would love more people to talk about it because it's something that when I started looking into the economics of the AI space, like I always, the first question I asked myself is where's the money, right?

Juliette: Well, you know, where is the money? Okay, so what we can see at the moment is there's lots and lots of money in, you know, basically having chips and And selling chips. Okay, so NVIDIA stocks going through the roof. It's not clear to me. Um, [00:38:00] the large language models, some of them are making money. Some of them, you know, aren't making money.

Juliette: And then the application layer, which should come after the large language models. It's not clear to me that any of them are, as you say, Okay. anything much more than a wrapper around a large language model, which means that, you know, basically all that happens is the large language model just suck up a whole load of information from the application layer.

Juliette: So Jasper, I think for me, would be a really good example of that. So Jasper was a writing tool. And it's just not much more than You know, as you say, a wrapper around, you know, chat, GTP, and then chat, GTP offers that service. And then suddenly Jasper is completely out of business. So I think, I think there is a really, really interesting question, which is at some point in order to make this.

Juliette: Uh, you know, interesting for application builders, and I would class myself as an application builder. Um, you know, you do need to be able to add a lot more to the large language model, you know, than [00:39:00] just a wrapper. Um, and I think for me where that Comes from is a really deep understanding of a vertical industry.

Juliette: So that can either be an applied problem within a corporate and, you know, to, to some extent you could just call that automation, right? And automation, you know, everybody is trying to automate different things within their own companies. But if you have a really, really focused piece of vertical industry understanding, as long as.

Juliette: It's complex enough that somebody like a large language model wouldn't really want to get into the weeds of it. Um, then I think, I think, I think you can protect, you know, the application can be delivered and it can be built for us because we have, you know, a kind of, Almost like a lifetime of understanding of how people learn a lot of kind of psychology behind learning a lot of psychology about habit forming a lot of psychology [00:40:00] about life design a lot of you know that's what we're trying to bring to Uh, the prompt engineering in and around, you know, the AI, and hopefully that will be enough.

Juliette: My prediction. Um, and obviously anyone can have a prediction, right? I mean, I'm not saying it'll be true, but my prediction is, is that you will end up with large language models and there'll be a few of them. You don't need that many of them. As I said, we're hopping around from large language model to large language model, you know, depending on how cheap and how good it is.

Juliette: So when new ones come out, we'll just move the prompt engineering to a brand new large language model. Um, so my prediction is, is that, um, what will happen is, is that you will have, continue to have the large language models, and then you'll have these small specialist, uh, you know, language models, like hopefully, you know, uh, we will be producing an iridescent after Zavmo, so that we're trying to produce a large language model called Grandjan, um, so that is basically where we're going, that will be focused on [00:41:00] this specific idea, yeah.

Juliette: Yeah. of ensuring that any time you hit a new piece of content, it is delivered in a way that matches your neurodiversity needs. Sorry, that was one thing that I didn't really, uh, so one of the things that we're trying to do is in delivering it as well as ensure that if you are dyslexic or you do have ADHD or you do have, you know, uh, visual challenges or any, any stuff like that, that it will be displayed in the way that you want it to be.

Juliette: And so, you know, our niche is. Is this idea of kind of delivering new information to the brain in a way that, you know, is habit forming and highly you know, highly engaging. So that that's the way I would would see it. But I would love to hear from more entrepreneurs as to why they're not building more really interesting applications on top of large language models.

Juliette: Cause I think there is, as you say, I think you've identified a real problem in the value chain stack, you know, which is, you know, where is, where are [00:42:00] these applications? Cause I don't really see that many. I see exactly what you're saying.

Mehmet: Yeah, because I have a theory because You know, and I can't blame them, at least in, in, in the geography where I live, because a lot of people, they think is okay.

Mehmet: Let me hop on a bandwagon, right? So, so now it's AI. Okay. How is the easiest way I can claim that I have an AI powered, you know, You know, uh, company or software, whatever. It's just, I use the API of any of the big three, four people who provide this. And then, you know, but I tell people like, Hey, you can do that.

Mehmet: But guess what? Two things to your point, Juliette, you mentioned this. About the example of jasper after a couple of months they would decide that they will add this feature that you thought like it's cool And then you are busted you're out of business or the second thing that might happen is that you have a kind of a [00:43:00] Single point of failure.

Mehmet: I like to call it because you are relying only and only on that LLM and for any reason, if they decide and don't get me wrong people and I have respect for this company for Google, but Google there, for example, very famous of killing projects, right? So, so they can kill, for example, Gemini, like who knows, let's say, so you need to put this in mind and I'm, I'm You know, and you mentioned it, Juliettete, like you're keeping discovering which LLM, you know, can provide you exactly what you do.

Mehmet: And then you're going to build your own one in the future. So you have this, I like, you know, this kind of vision and all it comes from something you started with. It's, it's a pain. It's a problem that you are trying to solve. And I think this is a debate Juliette really between entrepreneurs currently, you know, and maybe it's an old debate.

Mehmet: All of us, we know about it. Are you building a painkiller or are you building an, a vitamin product? Right. And, and just, you know, and I, now [00:44:00] I call it a cosmetic product. It's not even a vitamin because you're just trying to put some makeup that it has AI to just. You know, show it to a VC or an investor that, Hey, I have an AI company.

Mehmet: Do you want to invest with me? Like, this is my two cents only. You have

Juliette: to promise me that you will give me a clip of what you've just said, because I want that clip. I absolutely want that clip of what you've just said about vitamin pills and, you know, true pain. Because I think you've literally just summed up the problem that the value chain of AI has.

Juliette: It is exactly that. It is that people are really struggling to Find application builders, you know, uh, the, the other thing that I think is really interested about what you said is to be honest with you, it's, it's the story of entrepreneurship holistically. It's not just, uh, AI, and I'm sure now thinking about it, you probably meant it, you know, more globally than, than AI, [00:45:00] that there are a lot of me too.

Juliette: Companies. They are a lot of companies that really aren't focused on pain. They're focused on building, you know, making money or copying another idea or, and I, and I think. You know, it would be great to get back to the idea that entrepreneurs built products that, you know, really did solve pay it pain. And because of the pain that they're solving, you know, they were able to, you know, deliver value and get value from that.

Juliette: You know, I think it's a tough old road for entrepreneurs. You know, if I think about my journey in growth engineering, you know, Rightly or wrongly, we never took VC money. You know, we basically just did it the old fashioned hard way because we were solving pain and because we were solving pain, we thought, you know, we could compete, but then what we, I mean, we, we can compete, but we've never got as big as some of our competitors who have far worse products.

Juliette: It's just that they were given millions and millions of dollars, you know, to basically. Push into a product that is fundamentally mediocre, you know, and you just think, [00:46:00] well, it would be better if we could just focus on, you know, eradicating world problems rather than covering up those world problems with a load of money and potentially, you know, just, you know, creating products which aren't as you say, just vitamin pills, just vaporware, just they're not real things.

Juliette: It's just they've had so much money pumped into them. It's, it's, it's not, it's not really kind of making the world a better place. I mean, I love that. That's, I'm going to be thinking about that for a long time, but yeah, so, so interesting. And I'm, you must, you must send me that quote. I mean, obviously if you,

Mehmet: yeah.

Mehmet: So, so, so just, just one thing, I also learned the hard way Juliette and maybe, you know, I rarely talk. Oh, more than my guests on the show, but just because you mentioned this. So for me at some stage, I thought, you know, like I was fascinated. Oh, like look at these guys. They started their company out of [00:47:00] Silicon Valley or out of London or out of Dubai or out of Singapore, whatever.

Mehmet: And then, Oh, look like how much money they did. And, you know, I was young, of course, at that time. And okay. I had the itch. I need to build something that Makes me money. And then I discovered that actually we have, and this is one of the things I regret that I should have thought the other way around how I can impact, you know, people's lives, solving their pain, their problems, whether it's like personal business, whatever pain, any pain.

Mehmet: And then actually, you know what, like once you start to solve the problem, the money will come afterwards. Like you will be rewarded, you know, after that. And I think this is the, you know, the Holy grail of, of the, of the entrepreneurship. Journey in, in total, starting with a, with a end of mind. I, I, I like to call it, of course.

Mehmet: And regarding what you asked me for, this is recorded, Juliette, of course, but I gotta send you a snippet of that now as we are almost coming to an end. Tell me a little bit more about the book. You know, what is it about, and, you know, [00:48:00] when is it expected to be?

Juliette: So, so, so, so the book is really a combination of, of thoughts with regards to learning theory and how technology can be applied to what we know about how the brain works.

Juliette: So again, going back to your really important point about, you know, are we solving pain? So let's paint a picture of how Well, let's imagine what a classroom is like, you know, you might have been the kid in the classroom like me, you know, completely confused about what was going on most of the time and then, you know, in my report, it was like, could do better.

Juliette: Lack of focus doesn't concentrate, yada, yada, yada. But then I'm the person that, you know, I don't know, might go, go up to a lecturer or something afterwards and have a really fascinating conversation and being able to engage at a really high level because I'm good at conversation. And I'm, I'm, so I'm not unintelligent.

Juliette: I'm just not able to. sit [00:49:00] still in a classroom. I mean, I still struggle, you know, to sit still. I'm still kind of all over like that. I still have a short term memory. I still have all of those issues that you have when you're. So what, what I'm trying to convey in the book is, is look, we really need to give some thought as to how we are educating ourselves.

Juliette: We know how quickly, you know, AI can take on new information. AI can take on new information in 10 minutes. You know, in the same space, it would take a human to learn every single day of its life if it died at 85. Okay, so we are talking really quick, be able to assimilate information. So human beings have to be able to stay curious and stay, you know, interested in, in, in, in their future.

Juliette: And so the book is really about saying, Hey, look, this is the way the brain works. It's miraculous. It's, you know, it's magical. It's really interesting. We know. how humans can gravitate towards curiosity that we know how they can be engaged. So how can we harness [00:50:00] this thing, which just, it's just a data machine.

Juliette: It just churns, you know, information out. And how can we apply that to our education system in order to ensure that humans can live their best life? You know, I mean, AI should serve, you know, the human race in order for the human race. to basically, you know, have, have the best, the best life, the optimal, most optimal life.

Juliette: And we know that when the brain is firing on all syndromes, we know when it's curious, we know when it's engaged, we know that humans are in flow. And when we are in flow, we love our lives. So let's get AI to basically help us get there. So that's what the book is, is, is basically about. Um, the learning revolution it's called.

Juliette: Um, and yeah, I will, um, I will drop you a DM when it, when it lands, but it should be in and around sort of September time.

Mehmet: Fantastic. Final thing where people can get in touch and, you know, [00:51:00] follow what you do.

Juliette: So, yeah, best places. LinkedIn. Um, I am big on LinkedIn. I'm, uh, just in the process. Everyone keeps hounding me, Juliette, will you please do a day in the life diary on LinkedIn video video vlog?

Juliette: So you heard it here. It's happening. I've got to do it 10 minutes every day on or five minutes every day on on LinkedIn as a, as a vlog. So yeah, LinkedIn is the best place. I love to hear from people. So, you know, if, you know, Anybody's interested in just chatting about the future. Um, I would obviously love to, uh, hear from them.

Juliette: Um, So, yeah,

Mehmet: great. I will make sure that I will keep your profile LinkedIn profile in the show notes. And of course the, uh, all the websites, uh, for your, uh, companies also Juliettete at the end, really, I would like to thank you a lot for your thought provoking, you know, ideas that you shared with us, your thought provoking way of thinking about [00:52:00] education, whether, you know, whether it's for, for children or after we become, you know, professionals and we work.

Mehmet: So really, I, I enjoyed the conversation, Juliettete, and I appreciate the time. I know like it's a busy time for you now because you know, it's, it's in the building phase and getting things out. So I appreciate you took the time to speak to me today. And this is how I end my episodes. Usually this is for the audience.

Mehmet: If you just discovered this. Podcast by luck. Thank you very much for passing by. I hope you enjoyed if you did So, please subscribe and share it with your friends and colleagues we're trying to reach as much people as possible and if you are one of the People who keep sending me their comments their suggestions And questions, please keep doing so I read all of them.

Mehmet: I take everything into consideration So thank you very much for all your support. And as I say, thank you very much for tuning in We'll be again very soon. Thank you. Bye. Bye