Video

Summary

In this episode, we spotlight Adam, the co-founder of Stick’Em, the innovative startup that just won the prestigious $1 million Hult Prize. He explains how his team develops a robotics kit that costs a mere fraction of traditional models to make quality STEAM education more accessible to children everywhere.

This conversation, in two parts, offers a rare before-and-after glimpse of a startup on the cusp of greatness. First, hear from Adam during the Hult Prize Accelerator where he emphasises the importance of STEAM education in fostering skills like creative and analytical thinking to prepare students for an AI-infused world. Then, Daniel reconnects with Adam following his win and discusses how Stick’Em is already leveraging AI to streamline startup operations and plan their global expansion.

Tune in for an interesting look at the intersection of affordable education, social impact, and cutting-edge artificial intelligence.

‍

Transcription

Daniel Emmerson 00:02

Welcome to Foundational Impact, a podcast series that focuses on education and artificial intelligence from a non profit perspective. My name is Daniel Emmerson and I'm the Executive Director of Good Future Foundation. A nonprofit whose mission is to equip educators to confidently prepare all students, regardless of their background, to benefit from and succeed in an AI infused world.

Daniel Emmerson 00:27

So thank you ever so much once again for being part of Foundational Impact, our podcast series for Good Future Foundation. Today we're exploring what different startups are doing with artificial intelligence, both in their planning and their production and in the projects that they're pitching as part of the Hult Prize. I'd love to know a little bit about you, who you are and what you're doing and why you're here at the Hult Prize today. 

Adam 00:52

Awesome.Thank you for having me. So I'm Adam, and at Stick’Em, our mission is to bring quality STEAM education to every child. Because right now, in the age of AI, right, if you look at the World Economic Forum's job report, then five top skills that are required in the next decade for some reason doesn't have AI in there. And then we ask ourselves, like the five top skills that they are looking for are critical thinking, creative thinking, analytical thinking, technology use, and there's one more. I forgot what the last one was, but. But then we asked ourselves, why isn't AI one of these factors? And then we realised that if we as humans are lacking these skills, AI is going to completely outmatch us. So that's why they felt that it was really important for people to actually have these skills so that they will not be outmatched by AI, but instead be able to leverage on AI to be able to do more with their lives. So that was quite an interesting realisation for me because over the past, my entire childhood was essentially built around STEAM education. And maybe a lot of people would think that, oh, STEAM education, isn't it just science, technology, engineering, arts and maths. But actually STEAM education is an interdisciplinary application of these different subjects that allows them to also explore critical thinking, problem solving skills, their ability to communicate with their teammates, to solve a certain task at hand. But then we realized that across the world, there were nearly a billion children who are growing up without access to quality STEAM education. And we asked ourselves why. And there are two main reasons behind this. One is that current STEAM education infrastructure is either too complex or too expensive for most public schools to use. And the second reason is that there are simply not enough teachers who can teach these things. Because if you can teach coding and robotics, most people won't be teaching in a public school. Yeah. So we started out to really fix that problem.

Daniel Emmerson 02:51

So how are you doing that? I noticed you've got a prototype here or a model of your work. I'd love to know more about Stikk.

Adam 02:59

Yeah. So this is a. If a typical student in a school were to be using Stick’Em, this is a robot that they will probably commonly build.

Daniel Emmerson 03:07

Oh, it's Stick’Em. I didn't see the M bit on the. There we go.

Adam 03:10

Stick’Em is our company.

Daniel Emmerson 03:11

Yeah, thank you.

Adam 03:13

And through this activity, they'll be able to experience all the things that I mentioned earlier and how we are kind of like making this a lot more accessible and solving. The two problems that I mentioned earlier is that we developed this Steam kit that costs 1/10 of traditional robotics kits while having similar or very similar robustness to what you can get out of most other Steam kits.

Daniel Emmerson 03:34

Wow.

Adam 03:35

And at the same time, we are providing teachers because the teachers who may not be familiar or have any background in ICT will be thinking, oh, how would I even use this in the first place? So we provided them with an online platform that provides them with ready made materials. So instead of spending hours and hours figuring out how they are going to teach this, they can just go to online platform, use the lessons that are already there, that are very self directed, and then spend the time that they save on actually focusing on the students development.

Daniel Emmerson 04:06

And these, I mean, that looks like chopsticks, I suppose. Yeah, it is, it is.

Adam 04:09

So that's one of the ways we are really reducing the cost of actually producing these things and also making it a lot more relatable for the students. So we're from Singapore and we are addressing a lot of the primary schools and secondary schools in the Southeast Asian region.

Daniel Emmerson 04:26

Wow.

Adam 04:26

Yeah. And they are all telling me that this really helps them simplify the technologies that they thought were way, way too complex for them in a way that it's so easy for them to understand. And it helps teachers develop the confidence to actually use this in classes.

Daniel Emmerson 04:40

Right.

Adam 04:41

Yeah.

Daniel Emmerson 04:42

And what about you from a personal perspective? You said that STEAM is a big part of your, your day to day. What was it about that that you thought, hey, we need to try and solve this problem or we need to try and fix this issue.

Adam 04:55

Yeah. So my background was that I started building robots ever since I was 6 years old. And I grew up with this naive understanding that, well, if anybody wants to build these kind of robots, if they want to go for robotics competitions, I thought everybody had access to it until lo and behold, after I graduated from high school, I was teaching coding and robotics in like an enrichment center. And then the students would come up to me and ask, hey teacher, can I bring the robot home? And then I would have to say no, because it's not theirs.

Daniel Emmerson 05:23

Right.

Adam 05:23

And then the moment they hear this, you see like something switching off inside. And then, and back then I was just like, yeah, this is just the way it is.

Daniel Emmerson 05:30

Right.

Adam 05:31

Until.

Daniel Emmerson 05:32

This is because of cost, right?

Adam 05:33

Yeah, it's because of cost. Like back then I was using Lego to teach. And can you guess how much a single LEGO Robotics hit costs?

Daniel Emmerson 05:39

I don't know.

Adam 05:40

Take a guess in US dollars.

Daniel Emmerson 05:42

Ah, $300.

Adam 05:44

Good. That's a good guess. The actual number is US$650. So can you imagine how hard is it for parents to even imagine just think about buying these things for their kids? Like it's inaccessible to such a huge part of the population. And that's what we're really trying to change.

Daniel Emmerson 05:59

And I'm really keen to find out more about AI in particular. You mentioned it at the beginning. Are you using AI when it comes to your business development planning or your administration or project planning or, or even your involvement here at the Hult Prize?

Adam 06:13

Yeah, I think AI is unavoidable. I think at this stage AI is being used in every facet of our product development, our business development, and especially with GPT5. The agent mode is really, really helpful. Like it's able to go through your Google Drive folders and pick out the best pictures for you to use on your social media. It can also draft out emails for you. There's so much they can do that increases our productivity in a way that we've never seen before. And we are still constantly learning how can we leverage more other forms of AI automation tools that really helps us increase our productivity, increase the rate at which we can get things done. But at the same time we need to ensure that the critical work, the things that are really involved, really needs the human component to it. So for example, one of the things that artificial intelligence right now is very poor at is designing really good lessons for the kids because in each different country or in each different context or even age groups, they perceive information very differently. So AI currently we haven't been able to get it to do that yet. It's very good at creating content. It's very good at saying, oh, you should probably go through this, this, this, this, this in your classes if you want to teach about this certain concept. But it doesn't tell you the important ways of actually engaging the kids, right? Yeah. So those are like the pros and cons of the limitations that we're currently facing with AI.

Daniel Emmerson 07:43

And when it comes to that level of exposure, I'm thinking not necessarily about subject specific content, but around skills that young people might need at school. When it comes to AI thinking from the perspective of your company, Right. When you're hiring someone, you want them to have a level of AI literacy, for sure. What more can schools be doing, do you think, to help build the skills that the young people need?

Adam 08:06

I think that's a really important question that schools also need to really understand. And this is from my experience. So I think when it comes to students using AI, we can kind of divide them into two different tranches. One is students that are using AI to do the thinking for them, and then two students who are leveraging on AI to help them think better. Right. So unfortunately, a lot of students are ending up in the first tranche. They are using it to help them write their essays out. They're using it to do their math questions for them. And it'll be wrong, but they wouldn't know because they haven't thought through it. And so this is definitely the wrong way to use AI, I feel. But it's also very hard to get students to say, okay, how can we use AI to actually better myself? It's a very foreign concept to them because I feel like in schools, for most students, homework is a chore, learning is a chore. It's like they can't escape it because they're being forced into schools. So how do we change that mindset is something that I think I don't have an answer to yet, but I really hope someone can figure that out really soon.

Daniel Emmerson 09:17

Is there something happening in schools in Singapore that you've noticed that's addressing the change in technological advancement as far as AI is concerned?

Adam 09:27

Let me think in my university. So I'm from Singapore University of Technology and Design, and ever since the advent of ChatGPT came out, my school has been pushing very strongly for the integration of AI into all our work. And the way they have been doing it is, I mean, of course I have my own opinions, but I think they're doing some things right and some things could be done much better. But what they have been doing it is, for example, when we are trying to ideate, it's they. They try to encourage us to have AI as a companion to help us ideate different ideas to bounce ideas off of, like how we would come up with, like, a list of ideas as a team, and then we'll summit this to the AI and it'll give us, give us feedback or give us a bit of an analysis about what it thinks, these ideas, or how the ideas could be improved. Yeah. Other things would be like, let's see. When it comes to academic work, of course they do not encourage that. It's not encouraged. They enforce that we do not use it to write the essays for us. But it does encourage us to use AI to help us kind of like, understand what is the best way to structure our essays. Where can we put in, like, more relevant codes or more relevant evidence in a way that we won't be able to do it on our own? Because there's no way we have read enough literature to be able to sieve out all these really good examples or quotes.

Daniel Emmerson 10:54

But AI can do that using that example. Have you been able to link that with the work that you're doing for particularly this project? Is it helping you to improve? Is it challenging you? Are you using it as a. As a tool to increase your capacity in any way?

Adam 11:14

I think the value that the students will get out of STEAM Education is very critical to helping them understand how they should be using AI. It is because if they understand that, if they have a better competence in critical thinking, analytical thinking, and all the other characteristics that we mentioned from the World Economic Forum, they are very in a very good position to leverage AI to help them, not replace them. And. And I think, yeah, like, if I think everybody should take a read at the World Economic Forums job report, it really puts this kind of very well. And that's why I feel STEAM Education is really important in this day and age. It doesn't. This inherently doesn't involve AI on its own, but it helps equip students with the skills to become, to stay relevant in the age of AI.

Daniel Emmerson 12:01

What a wonderful place to wrap up. Thank you so very much for being on Foundational Impact.

Adam 12:05

A real pleasure. It's my pleasure. Thank you so much.

Daniel Emmerson 12:07

And that was Daniel's conversation with Stick’Em since recording this. I'm excited to say that they have since won the Hult Prize and we have them here again. In this next conversation, Daniel speaks to them about what's next. Let's get into this conversation right now. Okay. Adam, thank you so much once again for returning to Foundational Impact with very little time in between our last discussion and where we are today. You won the Hult Prize, which is phenomenal, an incredible achievement. So huge congratulations in achieving that I'm wondering if we can just pause for a moment before we jump into what that means for you as an organization and just reflect on that experience. Can you tell us about the day, what happened and how it felt to have won that?

Adam 13:06

Well, thanks, Daniel, for having me. Yeah, it's been a hot month, I think, since we last spoke, and winning the Hult Prize has changed a lot of things quite drastically in the way we see Stick’Em shaping out in the near future. What it essentially does, the prize that we won with the million dollars, is that it really jump-starts our ability to scale aggressively across multiple different countries. And it won't be just limited to the countries that we have now, which are mostly limited to Southeast Asian countries, but we'll start, we'll be able to start engaging regions in the South African continent as well as the American continent as well as Europe, where there are even in these developed countries, there are hundreds and thousands of schools that could still make very well use of STEAM education, affordable and intuitive STEAM education infrastructure.

Daniel Emmerson 14:05

So let's maybe think about that conversation that we had when we were looking at how we were looking at scaling and we were talking about AI and we were talking about the importance of AI in schools, but also as a business for you, I'm wondering if you've had a chance to reflect on how you might be bringing the capabilities particularly of agentic AI to your organization. Is it helping you with managing workload, with tackling administrative planning tasks, project management? Is there anything in that space that you're bringing AI to assist with?

Adam 14:49

Yeah. So currently recently ChatGPT released their or announced their agent mode where there's also like your ability to link different tasks together. That kind of a platform would definitely help and we are really looking forward to it being like officially launched in Singapore to start using it, to start really automating a lot of the mundane processes that we are doing manually now. There are quite a few automation tools out there like Zapier or Anydan, and these are very useful tools. But if you can, if we could have them all consolidated in a single platform like ChatGPT, that will make life for people who are developing or working on these platforms much easier. It just makes life easy to start being able to automate and manage all these things in a single platform. So operationally, yeah, it's helping with planning, it's going to help with responding to emails, crafting really creative email responses, or trying to understand, have a lot of contextual understanding when having a conversation with someone in a, in a country that we may not be, we may not be very familiar with. And of course, when it comes to engaging people from all over the world, language used to be a massive barrier, but now because of AI, it is becoming a very minor, a minor challenge when it comes to communication.

Daniel Emmerson 16:17

So thinking then about automation a little more, I'm wondering if you could go into a bit more detail on how tools like Zapier are being used already, to automate tasks.

Adam 16:32

Yeah. So I would differentiate Zapier and AI as two different things. Zapier is really excellent at automating very specific, fixed and predefined tasks. So for example, if someone were to make a purchase on our online shop and then the stripe side would send that extra invoice or that those purchase details to Zapier and Zapier will take these details and these details are obviously very fixed. There's not much flexibility there. So Zapier is able to take these details and craft an email based on those details. And then of course, the email's content is also already fixed, it just fills in blank spaces. And then you'll send out those emails to the customers, for example. So that's. And of course with Zapier you're able to link it to like ChatGPT using API or other AI tools. But predominantly it's used for very fixed, very predefined tasks, automating those. But what ChatGPT does differently is it will be able to, for example, another workflow that we're currently trying to develop and automate would be, let's say a potential customer emails us a few questions like, hey, I heard about Stick’Em through this event or that you guys won the Hult Prize and I'd like to get to know more. And I'm planning to use this for my primary school from, from let's say, Indonesia. And we want to engage them in creative thinking development and critical thinking development. Now usually that would definitely involve me or my co founders to have to understand their context maybe and try to craft out a response that is very customized to that specific question, those specific questions. And tools like Zapier would never be able to do that without any AI integration. But now with ChatGPT agent mode, it's able to contextualize. Okay, I'm going to tell it to study all of the emails that we have sent previously, as well as taking into account some of these additional details, our catalog, our product offerings, our service offerings that we provide to you into in the form of documents or links and then it's going to. And then Craft appropriate emails for all the subsequent emails that comes into this inbox, and then you can either say, okay, just send it out automatically or send me a ping to make sure that I check through the email, make sure the response is all right, and then I'll send it myself. Yep. So this is just one of many, like really creative automated automation tools that we can automated processes that we can use AI with.

Daniel Emmerson 19:23

In terms of the background work that needs to happen for you to make that effective and for you to feel confident in the communication and also the data management side of that workflow. How much time is involved would you say to. To organize and pre prepare and pre program any of that, any of that content? Is that a big job in itself or are you able to find relatively quick and easy ways of getting that done?

Adam 19:53

So far, in my experience, fine tuning it to really work perfectly takes a long time. But getting something basic up and running, something that's kind of agreeable, it's kind of workable, takes less than an hour. Like for example, gathering all these relevant data to kind of like give it the contextual information I think is the most important part. And that is quite easily done. We already have documents, it already has direct access to our drives, our websites. So it's. That part's quite easy. But the fine tuning part is where it gets a bit more time consuming. I wouldn't say it's time consuming. I would say it takes experience as well. Because as you start receiving all these various kinds of emails and then it starts crafting out its own responses based on what it knows, you're going to. Some of the emails are going to be okay, some of the emails are not going to be okay because it may be because it's very specific. Like for example, if it asks, hey, do you have a specific curriculum that's tied to this lesson that my government wants, then maybe we do not have the information or we did not provide contextual information regarding that, then it's probably going to say, no, we do not have it. But maybe I know that we actually have it. So those are some nuances that needs to be fixed over time. But the fixing is quite easy because all we need to do is just.

Daniel Emmerson 21:19

Add in more data and that requires oversight. Right. You need someone, particularly if there's cultural context as well, that needs to be considered around curriculum implementation. I'm wondering what that looks like as a role right now. Can you tell us about your org setup in terms of personnel and who's involved in those processes?

Adam 21:42

So currently we do not have anyone specifically dedicated to creating all these tools? It's me and my co founders that are creating all these tools ourselves. But of course going forward we are actually with the enterprise, we are expanding our team rigorously and they will be involved in working with and improving these tools as well. So I see ourselves still starting to, as I see the founders still being the ones to create the tools initially and get it to a basic working level and then getting like for example, our sales reps or organize, our school success officers who will be tuning these based on their own geographical locations of interest or their own basically the target audiences. Because I realized that from country to country cultures are very different, especially in Southeast Asia where there's just so many different kind of cultures and of course government regulations and different educational norms.

Daniel Emmerson 22:43

So when you're looking at that expansion piece for your team, how highly are you going to be regarding AI competency or at the very least AI literacy?

Adam 22:57

I think the beauty of these modern tools like ChatGPT and Perplexity is that there's not a high level of specific like AI competency that's required. Like even just two years ago there was such a role as prompt engineering and I'm sure there are still such roles as well. But nowadays these AI tools have become so intuitive and so human, contextually aligned. I don't know if that's the way to phrase it, but it really understands. It has a much easier time understanding what we're trying to do than we used to have to deal with about three years ago, two years ago with ChatGPT3 or ChatGPT2. Yeah, it's improved tremendously. So right now I wouldn't say that we specifically look for whether they have specific AI competence, but we will definitely ask them. Have you used ChatGPT for whatever projects have you used it for automation, automotive tools? What have you used these tools for? What tools have you used?

Daniel Emmerson 24:07

Can I ask what you would be looking for in ideal answers to those questions? What do you want to see people are doing and experimenting with?

Adam 24:16

So I would love to see candidates who have done personal like automated tools for their own, like projects for their own email inboxes or whatever their personal interests are. I would also like to ideally see someone who has used it extensively to create content like for get up messages or their emails or like the way they have. Basically as long as they have a lot of experience using the platform, I think it's, I think it's yeah. Good enough for us.

Daniel Emmerson 24:51

And how about on the data Management side of things. Obviously, if you're using APIs and you're tapping into potentially sensitive data and information coming from purchases or coming through the website, how important is it that candidates understand what responsible use of this technology looks like?

Adam 25:15

Yeah, that's a question that we are figuring out ourselves too, because it has only been a few months since we started really looking seriously into automating our tools and using AI to help with this specific part of the process. Like previously it was used to generate content for our Academy materials or creating new lesson materials, but now we are trying to integrate it more on an operational level. Yeah, and when it comes to data privacy, right now we are mostly focused on, firstly, not gathering data we don't need at all, especially the sensitive ones, because if we don't need them, we shouldn't have them. And secondly, if there are like for example, personal contact details, names of people or the addresses, which is obviously details that we'll get through order links or orders we're trying to prevent access to. So, okay, how do I put this? So as far As I understand, ChatGPT can only be linked, or at least the desktop version can only be linked to one Gmail or one Google account at a time. And so within that Google account, we'll be restricting the kind of data that Google account has access to. So obviously we wouldn't be tying our main Google account that has everything to this GPT agent. We'll only be giving it access to another Google account that we have direct control over what kind of data access it has. So obviously this account wouldn't have access to our financials, this account wouldn't have access to our credentials, and so on. Yeah, and we, and as right now, we're still learning, but I think the way we're going to go forward is to create multiple of these accounts for different specific use cases so that the data that has access to is also very segmented. Because currently, as far as I know, I do not think there's a way to segment the data a single desktop ChatGPT account has access to. As long as, as long as you just give it access to your Google Drive.

Daniel Emmerson 27:23

I think those things are companies are having conversations around what that can look like for business. Right. Because that's an ideal use case at the moment. You do need to segment and fragment what information it has access to. And depending on what model you're using, you might have concerns around where the data is going and what's happening to it. But let's maybe take a step back now. It's really insightful, Adam, to have an idea of Stick’Em's set up and how you're thinking this through from an AI perspective. I'm wondering if we can take a step back to your education for a moment because of course that's our main focus as an organization. We're really interested in ensuring that teachers have. Have the resources they need to become more confident in adopting AI responsibly for the benefit of their students and also for their own professional development. Is there anything that you might take from these experiences you're having at the moment as a startup thinking through AI collaboration and strategy? Are there any particular skills that you would have wanted more, more opportunity to engage with as a student if you were say, in high school today or just starting out in your degree?

Adam 28:49

So the question is what kind of experiences would I have wanted more when it comes to AI when I was a student?

Daniel Emmerson 28:56

Exactly right.

Adam 28:58

I think I was very fortunate to be in a university that really encouraged the usage of AI tools and AI and really actually not encouraging forcing us at some point in time to implement AI in our processes. Like for example for our. And you can search it up. So I recently graduated from Singapore University of Technology and Design and the university's new direction that was, that started being implemented late last year, if I remember correctly, is really trying to equip every single one of our students to be able to fully leverage the capabilities of various AI tools. So for example, one, for one of our entrepreneurship, for one of our entrepreneurship programs, it was mandatory like it was a literal milestone to say okay, please support your ideation document with some of the ideation, some of the ideas that was also generated in partnership with AI tools of your choice. So then we'll upload like okay, this is, this is our input to ChatGPT and this is its output. It could be like text based or it could be like image based and so on. Even for writing essays, they really encourage us to of course not let it write the essay for us, but to help us support with a wider range of evidence, a wider range of research that is of course all backed up by citations and website links. And it has, it's becoming more and more. I think they're trying to implement it on a wider scale as AI technology continues to evolve. And I think that's very interesting. For example, now they're exploring 3D generation of, generation of three dimensional designs for different kinds of components that enter mechanical engineers need to design for their projects, which I would not have been able to dream of three years ago and now it's possible. Or even tools like Lovable, which is able to create applications that are pretty robust from text based inputs. So you would see students coming up with MVPs within days? No, not within days, within hours. And you see the kind of progress that students are able to make when it comes to exploring ideas and creating new innovations. Because our school is also very focused on encouraging the students to innovate something during their four years starting there.

Daniel Emmerson 31:34

So when you said forced earlier that that was some a mandatory requirement, that means that it was incorporated into an assessment framework or a mark scheme. Can you tell us a little bit about what that looked like?

Adam 31:51

So it varies from module to module because the professor overseeing that module has full say on how it's graded and how AI should be implemented. Let me give a few examples. So for some of the design modules where we are designing a specific concept to a problem that they presented to us this submission. So first they'll be looking at assessing the quality of, the quality of the AI generated ideas or outputs in the form. Like, okay, so yeah, it can get a little bit confusing, but how do. You access it is firstly, how do. You design your input? Like, what is your query? And how do you work? How do you work with, for example, in this case, for me, I use ChatGPT most often. How do you work with ChatGPT to improve on the ideas that they have constantly given you? So for example, a poorly graded one would be, okay, please help me generate five different potential ideas on how I could solve, how I could solve water wastage in my school campus. And then maybe it just spits out five ideas. And if I submit that, that's going to get maybe one or two marks or zero, because there has been no discussion, there is no iterative development. So the way our school encourages us to leverage on ChatGPT tools is to use it as like a companion, something that you can bounce ideas off, something that you can develop more, more with, just more with together. So a better graded one would be maybe, hey, can you help me generate a few specific concepts that might be applicable to helping me reduce water wastage in toilets in my school, for example? And then let's say it generates five different responses or five different avenues in which I can potentially reduce water wastage. And that would be okay, let's look at this concept of automating or changing. Manual.

Daniel Emmerson 33:55

Manual. Manual.

Adam 33:57

What do you call it? Flush. Yeah, manual flush to automated flush. And what are the costs and considerations for that? What are some of the technologies that might be able to implement to an existing manual flush. So is there a way to change manual flushes into automated flushes that we as students can create? And having that discussion and going through a few ideas and really developing on the idea would probably give us a better grade for that segment of the evaluation process.

Daniel Emmerson 34:28

Some fantastic examples there, and I think that's really helpful for teachers to have an idea around what that looks like in a practical setting. And it certainly put you in a very good standing and very good position with what you've achieved already. With Stick’Em having only just recently graduated, I think. Is that right?

Adam 34:51

I graduated June this year.

Daniel Emmerson 34:53

Amazing stuff and really amazing work to learn about the journey that you've been on so far, and I know that many of our listeners will be excited to see what happens next. I learned recently that Esther has already purchased some of your Stick’Em. Through. The website, so we're really following this, this journey with you and excited to see what happens next. But Adam, thank you so, so much for being with us today. It's a pleasure speaking with you and good luck for all that you have moving forward.

Adam 35:28

Well, thank you so much, Daniel for having me. Always enjoy our conversations together. Looking forward to the next one.

Speaker 4 35:34

Thank you for listening to this episode. Be sure to like and subscribe wherever you get your podcasts. And be sure to tune into the next episode where we'll be speaking to the other Holt Prize teams.

‍