Signals Through the Noise: The World AI Forgot

[00:00:04] This is the Crypto Hipster Podcast. This is not a traditional interview show. These are perspective-driven conversations with founders, builders, and independent creators shaping what comes next.

[00:00:26] We go beyond headlines, beyond hype, and beyond price to explore ownership, freedom, and opportunity in the digital economy, where builders talk freedom, not price. There's a lot of talk right now about AI transformation, especially in emerging markets.

[00:00:55] Africa and the Global South and all of it. The Global South is a very hot topic. And all that to me, all what I see sounds really, really great. But I'm not sure that all of it's actually surrounded and grounded in reality. Most conversations about AI start from, really start from the top and then trickle down. But this conversation, this conversation doesn't. This is, so let's set this up.

[00:01:24] So today we're going to talk about how people actually use technology day to day. And I have Monty and Mitha Bisi from OrionX here who are building Uhuru inside that reality. And I'm looking forward to this. So let's get into it. I want to start there, you know, before we even start, you know, to talk about your product, right? I want to talk about when people talk about AI transformation in Africa, why are they consistently getting wrong?

[00:01:53] And how do they use technology on a day to day basis? All right, that's a great question. I can jump in and I'll let Monty join in as well. So most people in Africa, they use feature phones, low data smartphones, and their primary digital channel is WhatsApp and SMS. That's where the chat with their family, you know, pay for their airtime, run side hustles, even help their kids study.

[00:02:21] See, the misunderstanding is thinking that digital transformation means replacing what exists with what is already there. But at Orion, like what me and Monty are building, we do the opposite of that. We meet people exactly where they are. So we're going to be doing that. Uhuru is Africa's first LLM that lives natively inside WhatsApp and lives in SMS. It also has a new mobile application that we're going to be doing. It doesn't need no data plan.

[00:02:49] It doesn't need any airtime per se. You just have to subscribe to your bundle and you're in and you can use it and consume all the content you need as and when you need it. So that's basically what we're trying to sort of bridge becoming the intelligence layer and infrastructure layer of Africa. Yeah. Yeah.

[00:03:08] And just to chime in there, you know, what's happening right now when you look across the continent, there is the transformation that we're seeing is transformation that's really driven by an agenda from outside the continent. Because when you look at the funding, where does the funding go for air startups? Currently, it's the hottest air startups in Africa are language models. So it's translation languages. That's what they're doing.

[00:03:35] And that's where the funding from, you know, the likes of the big tech companies. That's where the funding is going. You won't actually find an air startup that's has a large language model and is just trying to develop an air model that's specifically for the content that's generalized. So general, general large language model, you won't find it because it's going to languages. So it's really it's not to solve the problems of the content.

[00:04:03] It's actually because once you can speak the language, then you know that adoption will be high and that drives like profits, etc. But does it actually solve the problem? That's really the question. So what I hear you said is there is a major disconnect between what's happening on the ground and what outsiders are saying should be happening. Correct.

[00:04:31] How how do you get people to understand and that gap and how do you get them to help build on the ground? How do you close that gap? You can go ahead, Martin. Yes, that's an excellent question. You know, closing the gap is is something that we've been trying to figure out ourselves, you know, talking to customers, releasing product and just testing, etc.

[00:04:56] But what we're seeing now is we're trying to educate people. So really, when we talk to a lot of the average citizen, a lot of them aren't really don't have digital skills. I'll give an example. We trained over 300 teachers in Ghana earlier this year, and we found that some of them haven't even touched a computer in their life.

[00:05:22] You know, so they don't actually have the just digital skills, let alone the AI skills, you know, learning how to use the air. So they see this, they don't even know what AI is, because AI, when you ask the average citizen here in Botswana, they'll say it's artificial insemination. I got I got a lot of that in the workshop that we run, which is for to do with, you know, agriculture, nothing to do with AI, with technology.

[00:05:46] So it's actually just the basic of educating people first and train them on digital skills to try actually understand what is actually happening with this intelligence that's rolling out across the world. And we have and now the issues, the outside, the agenda is just releasing products without reducing products without educating people, because as long as they use the product, then it's fine.

[00:06:14] But if they don't get educated, that's fine. But if they don't get educated, then they don't know how to use the product to actually solve the problem that we're facing here. So I'll say that's how we're trying to close the gap. Yeah, and Jamil, like, like you mentioned earlier, when you said how the outside world is looking at us, they are assuming that we have consistent internet across the whole continent, right? And that's a very first world kind of situation. Data is the most expensive here in Botswana, in the whole of Africa, our country has the highest.

[00:06:44] Oh, no, we're number two in the continent. Yeah, okay. Number two in the world. In the world. Oh, wow. Well, we keep adding up on the on the on the stacks. But the thing is, the point is data is expensive. And that is our main consumption of the internet. We don't. Yes, this routers. Yes, this Wi Fi and now more recently, the Starlink.

[00:07:02] And yes, AI communication is aware, like there's an awareness of there is something called AI, what it does, how it works, maybe a misconception conception to different people as and when, because obviously, there's a lot of media around it. There's only like chat GBT, etc. Those are the names that they hear a lot of. But what I'm basically trying to say is, there's a perception that they're going to roll out these tools, they're going to be consumed by Africans, and they're going to be scaled.

[00:07:38] And then you will see a lot of the data that we use them, and they're going to be investing. And then you will see, if you can avoid that. And you will see, if you can avoid that, you can avoid that as well. bridge that we're really closing the web as well. I do want to I do want to get into your use of WhatsApp and SMS.

[00:08:05] But I, you know, but we saw you said so far, there's what I see is a disconnect, right? I'm sitting here in the US. Everybody's like AI knows everything about everybody. And it's an accurate representation of the world. And what you're saying is that in Africa, it's actually it's actually not it's actually people heard of AI, it's not being used. So the AI that I see here, you know, much different than AI that you see,

[00:08:34] you know, how is our world being misrepresented? And what solutions need to be in place in order to get a uniformity? Yeah, no, that's, that's a great question. So, Jamil, like, the thing is, the misrepresentation just comes from the use cases that Africans would want to use it for, right? Most people in the US are things like white collar jobs, you know, like accounting, finance, etc, like in office kind of situations, we

[00:09:03] have a lot of those here in in in Africa. But most times, it's more cattle, like we're farming, we're doing arguing, we're doing mining, we're like in that space, right? So our use cases for AI are more like asking general questions, how to taking photos of things like very light use AI, which I think, to be honest, and to be fair, globally, that's probably what most people are using AI for. What real people should be using AI for is for

[00:09:29] building, creating, analyzing, scaling, producing, manufacturing, you know, creating jobs, when they use case of AI, but the problem is, there isn't an education that earlier to my co-founders point, point of educating people. And say, okay, this is a problem. It's how you use cases, you can simply use it like this, this AI is used for this, this one is used for this, like, there's no categorization, there's like this overload, it's overload of

[00:09:59] information, or how this AI does this, this new air comes out in the next day, it's a new AI. And that already just makes people discouraged. And I'm sure that's also worth an issue as well, like, so many things that are coming, everyone's trying to, it's like, what's going on, let's slow down a little bit. And the biggest issue here in Africa is because jobs is a very big sector, and a very big pain point. And the way AI has been produced and introduced, it's, we're going to be optimized,

[00:10:28] efficient and cut off the workforce. Already, it's on a high unemployment space. Now we're pushing for something that's going to increase it. So why would I want to learn about something that's going to replace me or potentially destroy my, my livelihood, etc. So that's really one of the biggest subsistence. Well, I mean, our youth unemployment here is 40%. I mean, and the most advanced economy, South Africa, it's

[00:10:53] above 50%. So when you hear that something here that's threatening a job, or then government won't, won't be happy, private sector will be happy, citizens won't be happy. You know, if that's the case, but I think you bring up an important point and what so 4%, if you look at the global large language models, if you look at the African data in them, with the

[00:11:20] dataset that they've gotten, 4% is African data. So already is a misrepresentation of African data within the global large language model, which is less than 4% for a contract with 1.4 billion people. So and so already now the issues as well as the date, the information that is being produced with when you use these models is actually not the truth. You know, I've tested it myself in terms of I

[00:11:48] asked, we have a famous chief from my tribe, and I asked that who's the famous chief, and he gave me some random person's name, justify created a whole history, justified it and explained and gave it to me. And that's not it's not true. You know, I even vetted and asked the elders. So now I asked it, I told it, no, that's not true. That's not the who is it, then gave me another name and apologized for saying the previous one still

[00:12:17] not true. And you see, so because of that lack of information, which majority of it is manual here, it's either manual, like in the paper, or is it digital and not online? You know, so you'll find that it's difficult. That's why we're worried. I mean, I once found the health data for one of the regions here in the country on in a random government employees laptop, random, like not high

[00:12:44] level or low level, just random and use the only one who had the data. So it's scattered as well. So because of that, it creates problems of how do we train these models on African information when it's scattered, not digital? I want to talk about Uhuru because it seems like more people have access to WhatsApp and SMS. You're delivering it through that, right? You know, so you're not losing, you're not losing a standalone AI infrastructure app. So on the surface,

[00:13:11] it sounds like it's pragmatic, but I want to challenge it a bit. So are you actually building new infrastructure or are you adapting yourself to the limitations of what already exists? We are actually building new infrastructure. And the reason we decided to go that route is because the current one is actually not solving the problem. It's it's it's it's not solving the problem at all. And we actually need to for once,

[00:13:40] Africa needs to have a technology that actually solves our problems. We've for the past 50, 60 years have just been consuming Western Eastern technology and haven't produced much for ourselves, you know, and we've paid the consequences. You know, right now, you know, we're doing also doing specialized models and we started with biodiversity. You know, that's the hot topic at the moment as said by the World Economic Forum as well in Davos.

[00:14:09] And, you know, right now we have to pay, you know, American companies, German companies to access our genetic data from a biodiversity database that sits outside. So the data that you only find the species here on the continent as it has the most biodiversity variation in the world. But the data, it sits outside and we have to pay for it.

[00:14:39] You see, and you see, that doesn't make sense. How are we paying for something that's ours? You know, to access it. So, um, hence why we started with biodiversity going in there. So we had to think, okay, we just can't do AI models. We have to also do the cloud. We have to make sure it's sovereign and also the AI model, because this is we, we, for once, we need to be the drivers of our own technology and our own future.

[00:15:06] Because, you know, the future just doesn't happen. You know, we build the future. So for once, can we actually be the ones building the future? So that's, that's, that's, that's why we took that route. Yeah. So I was asking you, is it, is it, is, is, are you guys constrained by ceiling here? And, you know, when does that advance start being a ceiling? And then, you know, in real time, like revenue or, or investment, what's it costing you?

[00:15:37] Oh, yeah, no, it's, it's, it's a problem, the costing part. And I'll, I'll say why. Right now, Africa contributes less than 1% to the global compute. So that's the 1% is coming from Africa. Now, that means we have to rent these GPUs from outside, you know, in Europe, and that's, that's quite pricing, you know, so it's, it's, that it's hitting us hard.

[00:16:05] So we had to really be innovative in the way that we've designed the model so that we can keep our pricing low, uh, and make it still affordable for the average citizen. Um, and we had to be really creative with it, you know, cause, um, if you're gonna offer cloud and also offer an AI model, that's gonna come at a cost. But if you innovate and you're creative, you can actually design something that's still affordable for them.

[00:16:32] Uh, but you're not being hit hard on in the books in the, you know, in the money side. Um, yeah, so, but it's because of, you know, once we have these data centers or AI factories, as they call them here on the continent, it, the, our costing drives down because they are, it's right here. But at the moment, uh, we have to use the GPUs that are outside the continent. And that's, that's the issue.

[00:17:00] So I think there's only one at the moment. Strive. Strive is one of the Africa's billionaires. He's the only one at the moment doing the AI data center. Um, I want to find out then where a Hoover actually outperforms what exists today. Um, we're gonna be running a benchmark very soon.

[00:17:26] And so far when we did our internal benchmarks, we outperformed, uh, chat GPT and, um, Claude's Opus 4.6. And then just to give a point of clarity, it's not just WhatsApp and SMS, but that's just how we're like distributing it here. There is a web portal, just like your chat GPTs and your Gemini's that you can log into and engage with it traditionally.

[00:17:48] We also have like a Google docs, uh, it's called you docs, you, uh, you press, you sheets, everything like pretty much all those products on the cloud within a guru. So it's like this productivity app ecosystem where people will eventually be able to build websites from WhatsApp. It has a lot of agentic models so that you're able to like ask it to do certain things and it would install what is required, whatever inputs are required.

[00:18:15] So if I, if I, if I dropped a user and like, say I had, I had this rent, I have this user, I'd drop them in your platform. Right. What can they do better than what they could do today? And then what are they constrained by? Like what can't they do better? Anything that's good contextually around African internet, uh, conversation, you can get everything, anything like generating images that are very high quality. Um, I would say that our image model right now is better than chat GPTs. It may not be better than Gemini's, but it's definitely better than chat GPT's one.

[00:18:46] Um, coding, like I said, it's as good as 4.6 and 4.6 Claude is, well, now it's 4.7, but 4.6 Claude is considered to be industry leading. So I would say that we're part, yeah, well, yeah, we're part there. And then I just general responses. It's, that is, it's a hit and miss based on how you've asked the question. And like, it's, I wouldn't say that we're the best there, but we're definitely in the top three. Yeah.

[00:19:11] All right. So I want to shift a little bit because whenever people talk about Africa, they talk about global challenge. You talk about development, they talk about impact, all that. Right. So sometimes it sounds to me, it sounds, it sounds really good. Like too good, like, like positioning more than like reality. So that narrative that you hear at the world bank or whenever you hear, you know, how much of that narrative is, is real.

[00:19:38] And how much of it is necessary to attract funding, tension, partnerships, you know, all that. What's the difference? Oh yeah. That it's a, it's a, it's a complex one. Cause yes, there've been many cases where, uh, some founders will talk big just for the sake of funding. You know, we have over 2000 languages.

[00:20:02] So you can imagine the difficulty of documenting over 2000 languages in the continent. So, but I would say that that narrative is more of the past type thing. It's, it's not what you're seeing now. Now you're seeing genuine, um, young people really trying to solve the problems of the continent. Cause we have serious problems with serious problems and they need to be solved.

[00:20:28] So I wouldn't say at the moment right now, there are these false, false, uh, you know, startups. I would say right now they're legit. They are trying to solve genuine problems because when they speak, you can tell they understand the problem. Uh, you can tell them, uh, you can even nowadays, you can now test the product immediately when they speak. So it is, these false, uh, companies, uh, it's, it's very difficult to do that now because people want to test immediately.

[00:20:58] You know, there's no reason why you can't have an MVP, you know, uh, in today's world. So I would say the narrative, uh, it's a thing of the past, but unfortunately, uh, it's still what people think immediately when they think of Africa and them developing their own technology. It's it's the immediately start to doubt and try, um, interrogate and undermine. Yeah. And undermine. Yeah. Yeah. Yeah. Quite a bit. Uh, like quite a bit.

[00:21:24] Yeah. Cause like Jamil, to be very honest, like as much as we mentioned earlier, like saying in terms of like the penetration and the awareness of artificial intelligence, Africans are very high literate. Like they're like very extremely, like, well, like the educated, like they tend to go to school a lot. They love trying to study higher professions and they love the professional services. That's what they love the most. Like the finances, your laws, your engineering. They love that. Those are the space that they really strive for.

[00:21:50] Yeah. Doctors, doctors. So like, so in that regard, it's like, yes, they are super intelligent in that capacity, but that does not mean that they can't create. So the upcoming youth like us, we've been, I've been in the space for like nine years building software, like it's in multiple countries. We've been doing this for a while. So it's, it's one of those things where not only have we have a proven track record, there is evidence.

[00:22:16] You can go to the site, you can test it out yourself. You can do your own penetration testing. You can, everyone's open. It's an open product. It's on the public domain. So anyone could interrogate it further to say, indeed, was this scaffolded? Was this a form of existing technology? What, what's, how is it working? And you can, you can do that for yourselves.

[00:22:36] So to be very honest, the teams that are sprouting across Africa are far and in between. If you think about it, a lot of the Western companies, in fact, even outsource a lot of the technologies capabilities to Africans. Like, like there's Nigerian offices, there's Kenyan offices, there's Ugandan offices that are just building software for the West. And then the West just puts a markup on it and then builds further. So it's like truly the workforce, the ones that are building those things are in Africa.

[00:23:05] The people that are building those things, it's now, instead of it being resold to an, to the West, and then it's distributed, they're now being the distributor. So that's the only difference. And then that's where the challenge comes is maybe there's a bit of an upsetness to be like, why you guys do this? You guys are just distributing. This is the pipeline. We should be the ones distributing. So let's discredit you to maybe create some sort of like mistrust or misguided direction in that capacity. That's a potential. That's just speculation. I have no evidence of that. Yeah.

[00:23:33] Right. So I, I think, I think you do have evidence though. I think that, you know, I want to know what was improved because like you mentioned founders, I want to know what's improved because, because of who exists, you know, as far as impact, what has, what is better now because you exist. I think that's a lot, but yeah, how much you can go in here. I'll take the education and you can take the, the, the, the, the, the, the, the, the, the, the impact.

[00:23:57] So because of who exists this, as I had mentioned earlier, provides almost something called, we call it an information explosion. It's like a bomb of information that most people who could not have direct access to AI now do. Right. That's something that does not exist. Like people in the villages and the outskirts everywhere, they can just consume AI quality content. Yes. Maybe it's just a text based version of it and not like the image and all that stuff.

[00:24:22] However, that's what we want. We want information, right? Like that's the seek, that's the ask for IT information technology. That's what it stands for. So the core of any technology is the information distribution and dissemination. And the biggest issues, the biggest game changer was partnering with mobile network operators. And they are helping us distribute these products at scale via text and people able to just download and subscribe and just work off it immediately.

[00:24:49] So in terms of day-to-day life, I would say that that is a big change for us. Like that helps a lot of people being able to consume and understand and ask questions. And then on the education space, we're doing some really great things. We've got a product called Grey Ed that Monty's going to jump into. It's doing some amazing stuff in Ghana specifically, as well as we've been in North and South America as well. Yeah. Yeah. So with the education part bit, so we have Grey Ed.

[00:25:17] So I'm dyslexic, right? So I struggled growing up. So I created Grey Ed a few years ago. And then I came and it was to really hyper-personalize education because, you know, I spent four hours a week with my tutor who specializes in dyslexia.

[00:25:37] And then I spent seven hours a day at school. But my language processing improved far greater in those four hours a week with him than with my hours a day at school. So that showed me, OK, when you personalize the education experience for a student, they can take in more in less time. So that's what we're doing with Grey Ed. So we actually look at the personality of the student. And then now teachers, they spend like, you know, I mean, teachers here spend two days on lesson planning, but two full days.

[00:26:06] And that's a time consuming. So what we did is we create also assist teachers with personalizing the lesson plan, but also cutting it down by 98% in terms of the time spent there. So we rolled out Grey Ed and, you know, when in Ghana went across 22 schools in Ghana, we're training teachers as well. We're trained 300. There's another 400 to be trained as well. All powered by Uhuru, by the way. This is all powered by Uhuru. Yeah, it's all powered by Uhuru. Yeah.

[00:26:37] And then in South Africa, we're rolling out across multiple schools there. Primary school, primary schools are starting there. We're even now about to partner with another NGO based in South Africa, where now we're going to, it's going to be close to 100 schools in South Africa across the provinces. And then here in Botswana, there's a specific school which we like using because it had the biggest impact there. Their pass rate was 61% in 2023.

[00:27:07] And then in 2024, their pass rate was 67%. And we joined them. So they take their examination end of year, December. So we joined them in October, 2024. So their pass rate in 2025 was 83%. So a huge jump between the previous year. And this is about 60%. Yeah, from 67 to 83. Yeah. Yeah. You know, 16% jump.

[00:27:33] That's in one year of implementation at low cost. Because there was a public school in a rural village that only got internet at the school last October, by the way. That's when they actually only, so they've been using WhatsApp or going to the Okutla, which is like where the, you know, town hall would be the equivalent, where there's free internet. So they would go there and make their lesson planning and everything.

[00:28:00] So that's been the hugest impact across education. Jump, helping public schools, whether it's rural, urban areas jump. Because our failure rate for the past 10 years is at 70% average for high school. So we're trying to flip that, you know, it's, it's way too high. So it sounds like the impact is real and it's really good. Yeah.

[00:28:27] So, you know, what I found that I'm glad you're, I'm glad you're working on it. And what I found though, is that if it's easy, everybody's going to do it. Right? Yeah. And so you, there had, you had, you had to have, you had to have faced constraints and challenges. And so what's the biggest, you know, constraint that you faced and how did you battle test that constraint? And then what you did when, when you did it, what almost broke, what made you kept going too?

[00:28:57] Honestly, and I think Monty will also concur with us is our, uh, GPUs. Oh my God, those things, those things are expensive as I have no idea. Like it is just so expensive and like the billing you on like a minute basis. But, um, at the time and still right now, there isn't a viable solution.

[00:29:18] And as our co-founder mentioned earlier, strive with Cassava is trying to set up in Africa to partner with Nvidia and roll out more affordable GPUs for, uh, LLM training and sovereign, sovereign AI. Yeah. Yeah. I would say that's, that's one of our biggest, I mean that, oh, and the data collection. Oh yes. Of course. Another big one. Yeah. Those are the two, those are the driver cost drivers. Yeah.

[00:29:48] Because this data is manual remember. So yeah, you have to be there physically to collect it, you know, these are many boxes. Yeah. You know, you have to get, and then you have to digitize it. And, you know, it's not the same as, you know, writing a script to scrape the internet and, you know, and can. Yeah. A Python script. That will just do it and you leave and you come back later and all your stuff is done. It's not. Yeah. No, it's not that at all. This is, oh, we need logistics and everything to collect all this data.

[00:30:18] You know, fight with government officials in regards to accessing the data, you know, so it's, it's, it's quite a bit. Yeah. Yeah. The data collection in the, in the, in the GPUs compute. Yeah. Compute cost. Yeah. I think those two things could stop anyone. It's like, it's like, it's like setting up a bank and then trying to put ATMs up. You know, like the cost of that is just like, holy sand.

[00:30:42] Like I have to get a network of ATMs across this region and has to cover a very good portion of it. So we're like, we need to get enough data that covers a good portion of use cases so that anyone who asks, it's giving you accurate data and making it more relevant than what you'd get on a chat GPT. So those are the kind of like, it's, it's expensive. I'll just tell you. You don't understand if someone stopped, like it would make sense. You know, you're calling. Yeah. If someone gave up, I'll be like, I understand. Yeah. Yeah. There's the same person here. Yeah.

[00:31:12] Why don't, so why don't you stop? It's our passion. It's, it's our, it's our calling. It's our calling to be honest. We were trying to impact a billion lives so that we have to. And we have no choice. We have no choice. We're at risk of being digitally colonized, you know, like a huge risk, you know, across the area. I mean, that's why biodiversity as well. We started with the specialized model there because they were very close to being fully digitally colonized in terms of even just control.

[00:31:42] You know, if the GPUs sit outside their data centers and we're here and we're using them. If they turn off the data center there, um, or for us, then we have no, nothing. We can do. Yeah. We don't have any, uh, GPUs to use. You know, so what it's like having your power station, the nation's power station in another continent and you don't own it, but it's giving you electricity.

[00:32:11] So you're not in control, you know, they can turn off whenever they want. And anytime they want a strong army, they can. Yeah. If they turn off and they want something from me, you can't fight, they'll just turn off. So because of that, you know, our long term plan is also to have, uh, AI data centers, you know, convert these, uh, data centers here to AI data centers. Cause we have no choice, you know, it's, uh, it's, it's at risk of being controlled. Hmm. So it sounds like you worked really hard.

[00:32:40] Um, yes. So you have, you worked on the distribution, but you, what you want is the adoption. And as we know, distribution and adoption are not the same thing. Right. So I would think that you want people, you want people to be repeat users. And so instead of just one time users say, Hey, we're, Hey, I tried it. You know, we want people to be repeated users and have retention and everything. So how do you get them? How do you get them there? Make sure they're, they're repaying the retained users.

[00:33:10] Yeah. I know Monty has a great, uh, answer for this as well, but currently what we're doing, right. Is there's a lot of AB testing that we try out, like, you know, how the AI responds, like, is it quirky? What memory looks like? Like we add, like we tweak and tailor different news products. Like, okay, this is working. This isn't working. But what we find is if it's useful, people will stay.

[00:33:34] Let me give you context right now, as much as it may not be wide news, but because we are like on the top of it, a lot of people are leaving chat GPT to go to like Claude or any other tool. Like a lot of people like in droves, but it's not being communicated or mentioned. Um, and, and that's, it's, it can easily disappear like a Nokia. It can disappear like a Kodak in the sense that just because it was leading doesn't mean it will always be number one. Right?

[00:34:00] Like if you produce and provide consistent great quality and steadiness and predictability, people are going to stay there. Like they're going to move. Like it's just what it is. Like if something comes along and it's great and it's consistently great, like for a period of time, the way this ecosystem and landscape is always shifting. Who knows? In the next three months, Uhuru will be number one across the planet. And everyone is using it for that period of time for whatever the use case is that it's, um, uh, propelling in.

[00:34:28] So those are the kinds of things that we're doing. And we're also working on something called Helios. It's obviously been announced. Um, I don't know if one of you would be okay with me sharing this, but basically what would happen is you'll be navigating in your website and it's like a, you know, when you're with IT support and you share your screen. And then it like helps you do stuff, it will be able to talk to you and be like, Oh no, click this or set this up. If you want your email set up, click here. Like it helps with a lot of boomers and like, um, older age people. It has a sort of assistant in your local language as well.

[00:34:57] And it can like, you can do anything pretty much. All these sites that you have to get on, like someone, some technical explanation gets sent by the AI saying, you need to install this driver and this driver and this driver. And you're like, Oh my God, I have to do these steps. And then you have to keep coming back. Now it's like, let me, let me follow you and help you do those steps. So I'm going to put that URL in you press the URL. Like, Oh, you see that bar there? Click that open this up, send this out just like that. And you go through it. So yeah.

[00:35:27] And it's important because the context is Africans are very skeptical people. Yeah. So I know that CROD has the computer use thing, but no, I know I can guarantee they're not going to be allowing it. Yeah. They're not going to let the computer. That's not going to happen. So it's at all. Already, they just don't trust digital things. So it having full control is not going to have that's why credit card payments are very low, low volume across the continent.

[00:35:54] Because people, they, the moment they input their credit card on their computer, they think it's gone. So yes, like now it's everyone has it. That's what they're thinking. So what I found is when you educate them, when you educate them, they actually stay. Because they did you as the person who is creating these things, you have all these ideas and everything that you like, I think they're going to use it. But now the user is seeing this thing for the first time and they don't even know what it is.

[00:36:22] They just know that if I say hi, it's going to say hi back. You know, that's, that's, and they're like, Oh, I can use this to write a report. No, you have to actually tell the user, you can use this to write, write a report. Then it's like, okay, but they all know now you can also do a business plan with it. So you have to tell them, or you can also do a business plan. Okay. Okay. But now they, they don't know that they can use it to, you know, make a tender document. So you have to tell them, or you can also do it. So it's educating them and tell them what they can actually do with this.

[00:36:50] Cause they, when they don't know, they won't think about it because, you know, they're not in the industry. So they don't actually know how the technology works to them. And then they just see this in this interface that I can talk to and it talks back to me and, you know, I tell it to do this and it actually gives it back to me. But in terms of now thinking, Oh, what could it possibly not do? And let me test it. They won't, they won't be thinking that. So it's about really educating the people on that. And yeah, we've found that that's, that's kept them here.

[00:37:19] And everybody talks about getting the next hundred million users, right? So let's, let's talk about like what that actually means. You know, I don't mean like technically, like, I mean, like, you know, so people like to talk about scale and, you know, what makes it, what has to be true for you guys to deserve to have access that level of reach?

[00:37:45] What, why, why would you have to look back and say, I deserve a hundred million users? What has to be true? I, to be honest, and this may be a controversial take. I don't think anyone deserves anything. Right. Like as in truly genuinely, I don't, I think. Um, it's all on merit. Like obviously if you're doing great things, people will follow and come, but also your location, your geographic location.

[00:38:11] Like if me and Monty were to do this exact same thing in the U S we would probably be a unicorn by like last week. Like it's just based on where we are right now. It's a lot more grinding gears, a lot more slower and turning, uh, the trust equity, African startup. Yeah. Like the African startup is a bit lower and it's like the scale of that is, is not like the best. So the, the, this micro things that create that resistance.

[00:38:36] So what you have to do as a founder that's in an environment like that is identify partners that can help you do that distribution. That scale, that first a hundred thousand in the, in the most minimal viable, uh, expense. Cause what we're trying to cut costs as startups in Africa, we always try to cut costs because capital is not readily available. There's not a VC sitting there saying, Hey, my son, I love you. Here's a million dollars. You know, like no, one's just throwing that at you. Like you, it is an extremely argue. Yeah.

[00:39:05] One company in the past five years received outside funding. And, and that funding was like 400,000 Bula, like, like $50,000. It wasn't like some huge amount of capital. Right. Yeah. And some of them have maybe, I think the highest was like a hundred thousand. I've never seen anything more than a hundred thousand dollars being issued to a bots company, but that's just the context. Right. And obviously our economy is smaller. We're 2 million people, but Singapore is about 2 point.

[00:39:33] And it's a small economy in terms of number of people, not in terms of GDP, but like they're able to graze way more than that. And that's because the capital in Africa is tighter fisted. It's where's your MVP? Where's your proof of content? By the time you're getting paid, you don't need to get paid. Like by the time they want to invest in you, you don't actually need the funding anymore. We literally build end to end systems with security, with proof of concepts, with GD and still don't get the funding as easily as the West do.

[00:40:01] And like, that is like, like, that's a very, very sticky issue in terms of like the scalability of a product to get that first 100,000 users. So we sweat for our users. We literally like, if you have 100,000 users in an African startup, man, like you'd have times 10, 20 or times 100 in a Western country with the same model. So, yeah. Yeah. Your customer acquisition cost is higher. Very, very high.

[00:40:30] You know, relative to outside because people just, I mean, just in our country, the, to get them to use technology is a mission and a half. It requires workshops and requires training. And then once they're on board, they do. So the, you know, in, in here, what spreads is those who are partnering with MNOs. That's what spreads. When you look at the startups that have really boomed across the country. Yeah. Yeah.

[00:40:59] All of them have MNO partners. One of them. All of them. It's not that word by mouth. It's very different landscape. So to, to, to get the 100 million users is, is you need to partner with the MNO because that's, that's been the formula here on the continent. And when you analyze how these startups have really, really scaled. And this is Flutterwave. This is Impessa. All of them. It's the partner of MNO in some shape or form.

[00:41:27] If you're an Impessa, if you're a Flutterwave, if you're any startup and you want to get the 100 million users in Africa, there's a 90% chance you need to get in with an MNO. I don't know any. I don't know any. Literally. I don't know any. There's none that have a hundred million users that are not with an MNO. Like none. I do not know any. I can't name them. Yeah.

[00:41:48] So if you and I are sitting here, you a year or two years from now, and I asked you, you know, what will have to have, have happened in order for the 100 million users. You have to, you would tell me we were, we partnered with MNOs. Yes. And yeah. And that's what we're doing. The, hence the data product being able to consume it and distribution is instant. Imagine every week we can guarantee you get an SMS about Uhuru. Like without us even being there. Like it's top of mind all the time.

[00:42:17] Like it costs the MNO nothing cause they're also getting an equity stake in it. It's, it's, it's just, it's a win-win. Yeah. So, so the hardest, the hardest, so the hardest truth that people that, that about Uhuru that people wouldn't say publicly. What, what would you say that is? The hardest truth. Yeah. Yeah. You won't get successful if you don't have a partner that can help you with distribution scale. Um, truly.

[00:42:47] And then like, yes, it could be an MNO, maybe even an ISP. Maybe if Starlink one day decides to be a virtual MNO and then it can distribute it via like, I don't know, like iMessenger messages or something like that. That could be a potential, but whoever, basically whoever connects connectivity on your phone. And that's our main consumption. Like the phone, whoever handles conversations or communication between devices, that is whatever it is, whether it's an MNO, whether it's a Starlink,

[00:43:15] whether it's a, a new form of technology with Bluetooth, whatever it is, whoever manages that and dominates this industry. Those are the people that distribution who has the largest number because. I mean, I'll, I'll give an example of like why they're important, you know, uh, for the first time ever Africa hosted the G20 last year, right? Um, in November and me and Thabu were invited to speak there. Mm-hmm.

[00:43:41] And we presented Uhuru and we presented it as this, uh, what, what we presented here. And the support that we got was tremendous because now we saw it as this tool that can solve our problems for the continent, but they saw it as this Pan-African technology. That's a, yeah. Yeah. That's there to really push the continent and give people, the ones who don't know how, the hows.

[00:44:11] And, uh, in order to have the fastest ever human development in history. That's what, that's how they saw it. Um, because they saw it as we get to keep our data, uh, within the continent, you know, 80% of African data is outside the continent right now. Now we get to keep it, keep it within. And they saw it as this intelligence that we get to spread to people, uh, and they can, you know, turn their villages into towns and their towns into cities. You know, that's how they saw it.

[00:44:39] And, you know, and it's built on homegrown data for the problems that are, uh, faced here on the continent.

[00:44:48] So now it got all that excitement and everything, but the, when you look at the, the traction, uh, the growth from there, it's not, it's, it, it, we did get a couple of, uh, um, you know, quite a number of new users, but it still doesn't match, uh, the WhatsApp growth. So it's very different environment here. And that's just the truth of the matter.

[00:45:15] We can't take Silicon Valley concepts and just plug them here. You know, you won't work. You'll raise the money, but you'll start up once or by five years. A hundred percent. And like if they, and if there is no MNO play or at least a distribution partner play. Some banking partners. Yeah. Your success rate is just lower. Uh, another thing is, uh, but I don't think a lot of VCs are interested in this.

[00:45:39] But if, um, there's a government play where the government says like, we're going to endorse this product and then mandate it to be utilized like that. I'm sure that works anyway, like in Starlink, Tesla, the, the rebates and things like it works anyway, but that would be another play. But, um, that's just a lot harder. Um, and MNO safe it's private, it's scalable and it's yeah, it's just very, very hard to get in. Like very, very hard. Like it is so hard. Months and months.

[00:46:09] Months and months. Like, like eight, eight plus months of just talking. And then another eight plus months of integration. Well, what you share with me today has been a fascinating journey and topic. And I think it's very interesting. And because what you shared is, is the part that nobody ever puts in the pitch deck. What stood out in this conversation wasn't the technology.

[00:46:37] It was the constraints. It was the constraints. Because when you're building an environments with limited infrastructure, expensive data and fragmented systems, you don't get to build for hype. You build for what actually works. And that changes everything.

[00:47:03] Because, you know, most, most of the AI conversation today is happening from the top down. But what we heard here is the opposite. Bottom up. Built around real usage, real limitations and real needs.

[00:47:27] And when you look at it that way, you start to realize something important. The future of this technology won't be decided by who has the best model. It will be decided by those who can make it usable in the hardest conditions.

[00:47:57] That's the real signal. That's the real signal.