A Little Bit About Juliette :

Juliette Denny is an entrepreneur and the founder of Iridescent Technologies. She holds a master’s degree from Harvard University. Juliette has over 25 years of experience in the learning technology space. She is passionate about using artificial intelligence to revolutionize education and make learning more personalized, engaging, and effective for all individuals. 

Through her company, Juliette aims to develop AI-powered learning platforms that can adapt to each learner’s unique needs, interests, and neurodiversity. When she’s not working on innovative edtech solutions, Juliette enjoys hiking in the mountains.

• Website (Growth Engineering) 
• Website (Iridescent Technology) 
• LinkedIn
• YouTube
• TikTok

TRANSCRIPTION:

JP McAvoy: Hi, thanks for joining us on today’s show. We’ve got Juliette Denny who is a Harvard post grad and building and learning AI partner. You’ll hear more about that. And she comes to us from Windsor, nestled in summer between Spain and France. Here’s my conversation with Juliette. 

Hi, thanks for joining us here today. Happy to have you on the show from Windsor, right?

Juliette Denny: Yeah, absolutely. The mountain Kingdom, beautiful.

JP McAvoy: That’s all I got for now, what brought you there originally?

Juliette Denny: We’ve got dogs, and we love hiking, and love being part of the outside. This place is just stunning. I can’t show you what I’m looking at, but it’s just literally mountains. And yeah, it’s just a piece of paradise, really.

JP McAvoy: That’s a good place, obviously, to start from as we begin today’s conversation. And obviously, you’re working on a number of different things, what’s occupying your days? Well, your most recent days, you’ve launched Iridescent, obviously, and you’re working on a few exciting projects.

Juliette Denny: Yeah, so we’re really just trying to just revolutionize the way we learn. Just a little thing that we’ve decided to just write a little thing. So I’ve been in the learning technology space for the last 25 years. And what we’re really interested in is trying to get people really curious about their personal development. So that’s what we were doing with growth engineering. Trying to gamify learning, trying to make it super social, but we kind of realized that there’s only so much that you can do without AI. So when AI came along, I was like, we’re gonna really do what we’ve always wanted to do with the learning experience, make it completely personal, make it all about the individual. And really, make it very conversational going back to the way Plato taught us about how we create curiosity and create lifelong learners. It really is about having that discussion and being able to situate the learning conversation away from one to many. More like, how do you learn? What do you want to learn? What’s gonna make you curious? Let’s deliver you that.

JP McAvoy: And as far as saying, you’re saying the AI is what’s empowering that, or allowing it to be personalized?

Juliette Denny: Totally, that’s what we’re doing. We’re coding up different AI agents to ensure that however you learn on whatever you need to be curious, the AI will be able to deliver. So let me give you an example of that. So when I was going through the school system, quite frankly, I was a nightmare. I didn’t really learn anything. I got chucked out of school. Not because I’m not smart and badass smart, but I just can’t remember anything. I have a very bad short term memory. And when I look at words that are longer than two letters, they just swim on the page. I can’t write,  it’s embarrassing. Even now, I get my Ws and all sorts of things mixed up. But I’m smart. I ended up getting my master’s from Harvard. But when I was having my formative years, I could have that learning delivered in a way that is perfect for dyslexics. So I’m talking about things like double spacing, I’m talking about key words being bolded. I’m talking about constant repetition because some dyslexic like myself have very short term, bad short term memories. I need a lot of reinforcement. I need a lot of repetition. This is the way, and I can move through content and learn through new things really quickly if I’m being delivered that learning in a way that really suits my neurodiversity needs. And so that is really the hub of what we’re trying to do. And we’re trying to make everything conversational because we know that in terms of learning pedagogy, delivering information, you just forget. We watch Netflix and then, what did you watch on TV last night? And that’s how the wine talks, that is literally the way our brains work. So what we need to do is we need to move that learning conversation further up so that we’re into evaluation into discussion into using that knowledge. And so this is all the fascinating, wonderful things we can do with AI. And this is what we’re working on in Iridescent Technology.

JP McAvoy: So many directions to go with what you just described there. First of all, brilliant and so needed, and that’s the power of the technology that’s allowing us to do these things now. So first place I want to go, or when you say you’re writing this, you’re like what? It sounds like such a daunting task. So what are you actually building the house specifically for? Is that actually occurring? Because it’s a lofty goal. So for each one of these, obviously, there’s a strategy to bring it to fruition.

Juliette Denny: Yeah. As you say, It is kind of 1% inspiration, 99% perspiration. But I truly do believe that you got to have the inspiration. We’ve got what we call the 4D process. So the 4D process is discovery. So we want the AI to talk to you and discover your future career path. So it is a conversation you have with the AI where the AI says, where are you now? What are you doing? Where do you want to go? So we are going on a voyage of discovery. We’re then going to discuss so the AI is going to go away, put together a development plan, and that development plan could be for years. If you’re currently a nurse and you want to become a doctor, or you’re currently in a call center and you want to be an engineer or whatever, or it could be much shorter. It could be, I’m currently a supervisor, there’s an opportunity for me to become a manager. I’m going to go years worth of management training in order to whatever it is, we’re going to discuss it. So we’ve made a discovery. The AI will then sit down and discuss with you what that training program looks like. You will commit to that training program, and that will go into your diary. So the second stage is discussing diarization. And then we go into delivery. So delivery is a little and often. It’s basically through the spectrum of whatever, however you want to learn. So the focus is very much a learning conversation. It might ask you a question, it might ask you to evaluate something, it might ask you to look at a video. 

So basically, it would be delivered out to you in line with your neurodiversity leads, in line with the type of learning assets you want. And then the most important stage arguably is putting that learning into action. So then it’s the demonstration stage where basically the AI will be creating scenarios for you to understand the impact of your new knowledge. So that is the learning pedagogy that we’re working towards. So that is the theory, and then what we’ve got is we’ve got multiple different AI agents that are looking at the quality of the content. So they’re looking at National Skills frameworks, they’re looking at National Qualifications frameworks to ensure that any of the content that is going to be delivered is in line with available credits. So we don’t want anybody putting in a lot of time, effort and energy in their personal development without knowing that they could effectively get a qualification from it. So that’s what is happening in the background, you’ve got a number of numerous AI agents all doing different things to ensure that you can really commit to your own personal development.

JP McAvoy: So how does somebody, I guess, put a lot of power in the hands of what’s been generated by the AI. So is it a challenge? Or do you think that people will have trouble trusting that the AI is choosing that path for the best place?

Juliette Denny: The thing is that the AI doesn’t do anything that’s not already in the public domain. All the AI says is, Juliette, I hear that you’re interested in AI. Okay, how interested are you in it? I’m like, very interested. Are you interested in the technical side? Are you interested in the conceptual side? And it’s gonna have conversations with me about what I want or where I’m interested in the AI ecosystem. Then it is gonna say, okay, Juliette, I’ve got a really clear idea that you’re interested in the future of AI. And conceptually, how would you go about building an AI platform? What I’m then going to do is I’m going to look at existing skills, frameworks, National Skills frameworks, and also recognise courses that will be on frameworks. I’m going to go to those frameworks because I know that there’s a quality standard there. I’m going to look at the learning objectives and the learning outcomes, and I’m going to pull together a course based on your needs. So the AI is not doing anything that is not being creative. If you like, the only thing it’s doing is allowing you to not worry about anything else other than what you want to do.

JP McAvoy: Yeah, it’s generative AI is what you’re saying. So where’s the datasets coming from? Where are the datasets that you’re employing?

Juliette Denny: For the launch, which is going to be in July. We are using Ofqual, which is a qualifications organization, national organization in the UK, and also the British Standards. So British Standards, which is to do with basically how you passport from one role to another. So this is in the UK, it is in the US. And as it is across Europe, they have a skills development framework which goes all the way up from when you’re about 14, 15 just entering into the workforce all the way up through to doctoral standard. And so they’ve mapped all the vocational professions to this framework. And so what we’re doing is we’re going to that framework saying, is it in existence? Okay, do you get credit for it? And if so, fine. We know what good looks like. And so we can deliver that content in line with those standards. So this is all in the open domain. The problem is that we just haven’t had the compute available to be able to take those brilliant frameworks and those ways of passporting joy.

JP McAvoy: As we’re having this discussion, we were talking about it. We now finally have the compute available to process to allow the AI to process, the level of, I guess, information is available. How are you meeting that challenge? And where’s the compute coming from as you’re building your company?

Juliette Denny: What we have seen with Google and working with AWS. We’ve got sort of a business consumer offer that we’re going to be pushing out every commerce for individuals, and we’re also launching with large corporations. I can’t say who they are now. But obviously, when we launch, I will be able to say who they are. And so depending on how often the large corporations want to bring in their own library. And so we’ve got options to basically post on either the Google Cloud or AWS, and that expands according to usage. One of the things with growth engineering, we will always bill clients on monthly fees. But moving forward with this kind of model, it’s very much going to be based on what’s happening on the server. So if you do have learners that are committed to using core AI because that’s what the learning companions called on a daily basis and interacting with it three hours a day, then obviously, they’re going to be paying more for that compute.

JP McAvoy:  Fair enough, makes complete sense. Now, as you start to reveal, I guess, what you can at this point driving towards the July launch, I think is where things are, as a lead up to, what can we expect to see in July?

Juliette Denny: In July, well, you should be able to go in and you should be able to go through the full 4D process. You should be able to create your own plan with regards to what your future looks like. You should be able to diarize, and then you should be able to engage in these learning conversations. What we’re really hoping is with the engagement and learning conversations that we can stop this one too many thing. So even if I’m learning about what conceptually AI is all about and how I can use that new knowledge to come up with ideas for businesses, you might be looking at AI from exactly the same framework you might want to. But because the way we learn is different, you could be presented with a different learning experience than me. And that’s the nub of what we’re trying to do. So we believe that there is a lot of information out there. YouTube’s full of it. The list goes on and on. There’s a lot of content, but content doesn’t deliver better learning outcomes. So what we’re concerned with is how you, as a learner, get the best learning experience possible so that you keep coming back to class to learn new things. Because really, if you think about what we can all achieve, what we can all achieve is predicated on how brilliant our brains are. So we want this learning experience to be literally brain free queued for your curiosity. So that’s what you should be able to experience. You should go to bed at night and go, I can’t wait for my next learning conversation.

JP McAvoy: That’s a great way to be. And so we were saying July where people find that if they’re interested in going through to it.

Juliette Denny: If you just go to iridescent.technology, that’s the website, you can sign up and meet us at the end of July. So we’re going to be sending out updates letting you know how we’re doing and all of this good stuff during the journey. But yeah, we need a set of beta testers. The more, the merrier. And also, the more obscure that your learning interests are, that’s also interesting because when we don’t have a library of content, that content is being made up around the learner. It’s completely personalized to the learner. The reality is you should be able to learn everything from how to be an expert fisherman, to how to be great a paddle, my partners doing that, yeah, all the way through to how AI can transform the learning landscape, which is what I’m into. So whatever you’re into, we should be able to deliver it through the platform.

JP McAvoy: Right, right. Well look forward to seeing that continue. And as we talk in July, see and actually go live, when we discuss that range of things that there are friends working on a project right now, you’d be interested here if there is any solution for it. I’ll be more specific here. I’ve got a friend that has a son that is communicative, he is autistic, and he’s been unable to really communicate with his mother for years. They’re just getting the spot now where they can actually communicate with each other. He’s maybe at 18 when they first, we’re able to begin doing this. And they’re now actually in a shape or form where they’re actually communicating with one another. And so we’re exploring the power of AI to facilitate that type of exercise as well. Is there anything that you’re building that would be something that could actually be employed towards that goal as well?

Juliette Denny: I am by no means saying that we are there now, but that is certainly something that we want to do. So we believe that the secret or kind of problem that we’re solving is not content, it is not qualifications. It is this idea that we should be able to make learning accessible. So as we move through this project, for example, when you look at how we’re delivering content out to dyslexics, it’s a pretty blunt instrument at the moment. But the more people that we can get onto the platform, and the more feedback we can get with regards to how we’re delivering that content to those with ADHD, autism, the more we are going to be able to tweak and improve the learning delivery. But we are in the early stages. I wouldn’t want to say that we could solve really complex cases. But do we understand that we have to, that’s what humans should be doing. We should be solving complex cases, and that’s certainly what we want to do. But even if we can move it slightly, even if we can not take the cases that are really difficult and really challenging, the individual humans have struggled with, but even if we can give AI some of those problems, so for example, if you’re learning literacy or you’re learning whatever it might be, there are lots of little changes that you can make very, very quickly in a individualized way through the power of AI that is very difficult for a teacher who’s got 60 people in their class. That’s just not possible for them to do in the same way as it’s not possible for standard e-learning to be able to move and shape the way it delivers the instructional process. It is one too many. You build it once, and you shoot it out to many people. So yeah, I’m not sure of that.

JP McAvoy: It does. And it’s interesting. Circling back to what you just said, we have to understand that we are early, this is all in its infancy, undulating with that vantage, or from that vantage point. What did things look like in, let’s say, two years from now?

Juliette Denny: I think one thing is very, very clear to me from working in learning development and education, the technology that underpins that in a corporate setting is that change happens at the rate of the slowest human. I think we can have the best AI in the world. And we do, arguably. We have literally got some really clever AI running around the place, and entrepreneurs like me are picking it out and trying to do interesting things with it. But it’s useless unless individuals actually decide they want to interact with it. So I would say that, coming back to the whole hype cycle thing, there is a lot of hype. But until individual people actually find it useful in their normal days, it’s not going to take over the world. And I’m not saying that as a bah humbug kind of like. It’s the reality. Most people still use Google for search, they don’t use AI. We might be in this world where everybody’s talking about AI, but a lot of people don’t want to use it. They’re not using it. So I think, yes. I can understand why everybody’s like it, it’s going to be a dramatically changed future. But the normal guy or the girl in the street, they’ve got to be convinced to change what they’re doing. And my question would be, are we introducing AI into people’s lives in a way that they actually want to use it? Because there are lots of people that I speak to when I have conversations about work at an AI first company, we’re trying to deliver something new that immediately turns around to, AI is scary. It could kill us, it could basically drive us to all our jobs. So there’s a big change management process here. And from my point of view, I think the best way to look at the emergence of AI as AI will become part of our daily lives as long as it’s useful to us. But I don’t think you’re ever going to get a situation where we’re forced to use AI and the world is taken over by AI even though we don’t want to use it. My learnt experience from change management, people move at the change they will feel comfortable with not just because technologists and your boss say you’ve got to do this.

JP McAvoy: Yeah. And I think what you’re talking about is when we get to that critical mass. And I think for a lot of these people that you’re describing, they may not even be aware to the extent to which they’re already using AI, and to which it already is part of daily society. And I think the pace at which that change is occurring continues to escalate. So I always ask the question from two years, because I think it really isn’t within that timeframe where we’ll have achieved that critical mass. And so then it becomes a question of the day of the user, or is there something, what are the friction points? What is it that’s required as opposed to doing a Google search to be using their GPT? And I don’t think it’s that far away to the point where we are all using it that way. A lot of people such as yourself are building some of the solutions to that problem to onboard or have people using it to the spot where it becomes ubiquitous. And I don’t think we’re too far away from that. I always ask the question for the people developing it where they see that in the spectrum because I am never too far away from here.

Juliette Denny: One of the things that I would say I find quite interesting is that most people use their mobile phones far more than they use their PCs. So we still haven’t got a decent AI solution on a mobile phone yet. Yeah, that is still somewhere else. So you can’t have a really powerful GPT on your phone. I would argue whether something like what we’re doing, you certainly wouldn’t be able to do the first two. So the discovery and the discussion, because there’s just too much going on. You might be able to do some elements of the delivery on your mobile phone. But I think that’s going to be a critical point once we get really good at AI solutions on your mobile phone. That will open up a whole other spectrum of people utilizing AI. But I think your point is really interesting because we all use AI in our daily world, it’s just we don’t know that we’re using it. A lot of the enterprise platforms we use, a lot of the time we’re interfacing with social media, for example, there will be components of AI that are running through that platform that you don’t know that you’re utilizing.

JP McAvoy: That’s certainly the case. And then all the questions or a lot of the thinking through some of the people that they’re doing is to discuss what is, going back to as you said, people are scared of it. So what are the guardrails? Or what guardrails can there be so that we can trust what is being generated? Do you have any sense of that? Do you have any ideas for yourself as you’re building what the guardrails are at this point?

Juliette Denny: For me, I think that it’s one of two things. First of all, I think that you as an individual need to choose how you interact with these AI companions. So that’s the way I view it. You’re gonna have different AI companions for different facets of your life. You might have a master AI, and then you might have a number of mini AIs, but the AI is basically going to be a reflection of how you interact with it. So the way I view what the future will be like with our AI companions is very much by interacting with the AI and by telling the AI what kind of experience you want to experience with that AI. That AI will take on the personality that you want it to take on so that it can be everything, anything from an executive assistant all the way through to just a search engine. So I believe that the guardrails should be very, very clear with the individual that is utilizing the AI. And I believe that you can have a very kind of linear, non invasive relationship with AI, if that’s the way you choose to have your relationship with AI. So I think there’s going to be a lot of upskilling for people to understand what information they give and what information they don’t want to give. Obviously, the more information you give it, the more useful it’s going to be. So for somebody like me, I think I would be willing to share a lot of my life, my interests, my passions, my career objectives, my soul with a digital AI in order to get an overall better experience of life. But I’m using the guardrails that should be with the individual. I think there’s a separate question with where all that data goes. 

And I personally believe that I would like the data to be held on my own server. So what I’d like is to use AI and these large language models to create my own personal assistant. And then I would like the data to be housed in the cloud, but it is locked for Juliette’s eyes only is the way I would like that to be. So that doesn’t mean that I can’t take advantage of the large language models. And the larger the ability for AI to be incredibly useful. But I don’t want my data to be used by anybody else other than me. I’d like to think that’s where we’re going. Because I do think that people worry about control. They worry about, oh, my goodness, what’s going to happen to these agents? Are they all going to gang up on me and then leave? Or whatever it might be, are they going to do nefarious things? Well, if you’ve gotten various things in your mind and you’re talking to your AI about nefarious things, then guess what? You’re gonna have a nefarious AI. But if you’re like me, unicorns, pink, glitter sparkles, then what is the AI going to be like? Well, it’s gonna be unicorn glitter sparkles, a bit of AI, a bit of Taylor Swift or whatever I’m into. So I think we need to take personal responsibility for coding up our own AI would be the way I would see the future.

JP McAvoy: I love that. You said so many great things within that. And one of the things that struck me was, as you said, the relationship with AI. People have to start getting used to this idea of that relationship, their relationship with AI. And as it’s built a whole slew of things, you say the security and a lot of implications that are being thought through right now. It’s all being developed right now. It’s actually all occurring right now in real time. And the interesting thought is you speak of agents ganging up foreseeable as well because I have had these types of discussions. There are those that are saying that we’re not that far away from. And I’ll throw this to you or see how you’d answer. If you’ve given it to yourself, including your soul, right? If you describe that, but it’s learned, and it’s the spot where it is now sent to you. So it’s interacting. Well, the same way an individual that you’re interacting with may try to persuade you or even manipulate you, it’s foreseeable that AI is going to be capable of doing the same thing in very short order. How do you respond to that?

Juliette Denny: I think it’s very much a discussion on the future of AI and how dark the future of AI can get. I think it is very much a reflection of how you see society. I don’t think there is a binary right or wrong answer. In life, you get people that are half full, and you get people who are kind of half empty. My personal view from actually being into the programming and prompt engineering of agents is that we are actually quite far away from a generalized intelligence that can itself, let’s say, create a company like I’m trying to create. I know that people talk about the fact that we are going to have AI that can come up with ideas that can basically hold all the other agents together to create a business, to create a company to end there won’t be anybody other than an autonomous AI within these companies. I personally think we’re quite far away because I think one of the things that we need to have is this idea of being able to look at multi dimensional subjects and being able to bring all of those together in order to create something new. I just don’t think AI is there. At that point, I think what AI is really good at doing is following instructions. I want you to do that, you will know when it’s complete. When it looks like this, you will follow the following quality criteria in order to understand whether you have completed the instruction to the quality that I want. Yes or no. And you will do this X amount of times. So AI is there doing that, but I don’t think AI is sitting there. I know AI is not sitting there going, oh, wouldn’t it be interesting if I basically got the color pink, got a unicorn and was really into motor racing. And basically, I was gonna create a super speedy pink unicorn that fostered glitter. It’s just not there yet. In the future, I’m sure. But right now, it’s not in a situation where it’s doing anything other than what humans tell it to do.

Juliette Denny: I love that. I love how you describe it. It’s not there yet, and we have to envision a future where that actually does exist. If somebody’s listening to this and just creating their own AI company right now, what are some of the first things, first principle is for them to be launching to be getting their business to up and running? And again, creating some of this future that we’re describing here today.

Juliette Denny: One of the first things is when you’re in the AI world, or when you’re slightly obsessed, like I am about learning about AI, you assume that everybody is actually using Chat GTP. And actually, they’re not. So the first thing that I would say is if you’re interested now start using it on a daily basis, and then you’ll suddenly find, like me, that you actually can’t get to any power without using AI, let alone anti working day. Then once you’ve started doing that, look at automation, look at how easy it is to automate really rudimentary tasks in order to make your life so much more efficient. And once you’ve done that, you’re kind of within the AI world. So it is very accessible for each individual person to really get a handle on what AI is, and what AI isn’t. But I think somebody said something about the world of AI in terms of, there are going to be those people that utilize AI and basically their lives to magnitude better. And those people that just stick their heads in the sand and don’t want to do anything with AI. And unfortunately, they will find themselves not necessarily left behind, but they’re not going to be able to have the advantages of what a more efficient, creative life looks like because I can take a lot of mundane tasks out of your life.

JP McAvoy: That’s another nice way of putting it. It isn’t here now, people are using it to a certain degree. It would be wise to become educated in how it’s currently being used. And think of the future, what things look like in the future. So Juliette, I really appreciate your time here today taking us through some of that. I look forward to watching the launch so please keep us in touch with that. If somebody is interested in connecting with you and what you’re about to launch there, what’s the best way to reach you?

Juliette Denny: I’m thinking about LinkedIn. I’m always chitter chattering and jabbering on LinkedIn so please do link me in, tag me. I’ve also a little Calendly link for virtual coffee. I love meeting new people, finding out what people think about things, especially in the world of AI. So yeah, be very, very happy too. I’m also on the other socials, but where I’m most active is LinkedIn. I love TikTok. So yeah, on Tik Tok as well, but just not quite as active as I am on LinkedIn.

JP McAvoy: So we’ll have all that in the show notes as well. Anybody who’s interested, please do look for that. And then Juliette, I love to end this show with one thing that someone listening can take with them through the rest of the day, through the rest of the week after the show is dropped. Is there something that’s worked for you as you’re building this company in life itself that when somebody listening may benefit from?

Juliette Denny: I think my kind of words of wisdom are summed up by Dale Carnegie when he says, people are most successful when they’re having fun at work. So I would say that the one thing that everyone should try to do is basically enjoy. Enjoy the ride. That would be my takeaway.

JP McAvoy: That’s right. Wonderful. Yes, thanks, just wonderful advice as it is. Enjoy it, enjoy it along the way. Be happy, spread the joy. Thanks so much for spreading your joy here today. We look forward to next time on The Millionaire’s Lawyer.