Industrial nations are competing for the lead on policy making, adoption and proprietors of Artificial intelligence solutions. The vibrations at the UAE at Ai Everything and France’s Ai Action Summit were reaching new heights.

While much of the global conversation on AI is centered around weaponisation and cyberattacks, many have taken a quiet lead on technological innovation especially in the US and in Asia.  Lagging behind is Africa.  The focus is just not on in the right place. African governments too concerned by “training” and “ethics” criticism from others have missed the boat on providing the required equipment for computing. “You cannot learn to drive a car without the vehicle!” says tech futurist Alex Tsado. If African nations want to compete for global wins, as a creator of AI-driven solutions, they can, with the conducive environment for infrastructure.

Despite these risks, there are promising AI initiatives emerging across Africa. Startups, policymakers, and academic institutions are working to harness AI for local needs, addressing challenges in healthcare, agriculture, education and financial inclusion. However, these efforts remain fragmented, and a lack of coherent, forward-thinking policies is slowing progress. While some governments have started drafting AI policies, many remain vague or underfunded, failing to create an enabling environment for AI innovation and investment.

Our AI Editor Faustine Ngila sat down with Alexander Tsado, a top voice advocating for Africa’s role in the current global AI revolution, to explore the AI challenges and opportunities for Africa.  Tsado has co-founded the Alliance for Africa’s Intelligence (AAI), a nonprofit dedicated to ensuring Africa becomes a key player in AI development rather than a passive adopter of foreign-built systems. He has also co-founded Ahura AI which boosts employee development, productivity, and engagement. 

This interview highlights the risks of the continent lagging behind due to inadequate policies, limited infrastructure, and insufficient government support. It examines the impact of various AI initiatives across Africa, shedding light on how startups, research institutions, and global tech firms are driving AI adoption despite regulatory and funding challenges. The discussion also delves into the policies being adopted by different African governments, assessing their effectiveness in fostering an enabling environment for AI innovation. 

Here’s the full interview excerpt, edited for clarity.

Let’s start by understanding your background in AI and your general insights about  the current AI revolution in the world.

I’ll start by saying I’m Nigerian, and grew up in the ancient city of Benin. Now, I spend a lot of my time in California. And I moved to California because of my job with NVIDIA. I was on the product team and I was responsible for working with all the cloud service providers to launch their first GPU instances. So this is the likes of AWS, Microsoft Azure, Google Cloud, and some of the other bigger cloud providers in Europe and in Asia. We worked with them to sell the idea of artificial intelligence and how that could potentially transform the world of innovation. 

The world has now transformed to a place where AI is almost synonymous with growth and innovation. And I left that role in 2020 and just about a year and a half before that, took on the heart of co-founding Alliance for AI, Alliance for Africa’s Intelligence, with the simple mission of ensuring Africa doesn’t get left behind in the age of destructive revolutions. 

The current level of diversity of the revolution we’re living through is artificial intelligence. But after this, there will be others like quantum computing, even blockchain. And so we recognize that it was really difficult for our systems. They were not set up in a way to keep up with all of this. Universities weren’t picking up and changing their curriculums fast enough. Neither were our high schools. Our governments were not set up to change fast enough. 

And so we decided that we will create from the philanthropic side an African-focused initiative that could create programs that would support our universities, support our governments and support our startups to keep up with innovation. And it’s pretty exciting what is coming. 

And based on your AI role at NVIDIA and building systems for AWS, Azure and others, what do you think would be the go-to-market strategies for cloud AI computing in Africa? 

Yes, I would have that expanded beyond cloud computing in Africa. I think the big deal is that we need AI, the age of AI to be different from the past age. In the past, Africa was mostly a consumer of innovation. 

Africa was a source of raw materials whereas the rest of the world would extract raw materials from Africa, take it outside, do the manufacturing, and then sell the finished product to Africans, sometimes 10 times more expensive, sometimes 100 times more expensive. In the age of AI, we must ensure that that is not the case. For several reasons. One is just bad economically. But the second is with AI, it’s actually worse in terms of harm. 

When you don’t build  the solutions that you use in AI, there’s a very big chance that the solution has, you know, there’s a high risk that it can harm you. Why is this? When you design AI solutions, there are several decisions you need to make from what data you use to train the AI model to how you interpret what you are seeing in the data. When you’re not the one designing the AI solution, you have blind spots. It’s not because you’re a bad person. You just have blind spots. 

You need to be aware of the way things affect those people and of historical issues that they have. For example, even within, even in the United States, where, you know, it was almost like ground zero for AI. There was a very poor, very bad history of, in the hiring space, where black people would not get hired for certain types of jobs. Women would not get hired for certain types of jobs, like engineering. 

And so when they build AI tools to become like the HR, to select people for jobs, it won’t give women jobs in engineering or think that black people were smart enough to get those kinds of jobs. That’s what happens when you build AI and you are not a black person and you’re not a woman. Very same thing happens in Africa with AI. And so that is why it’s extremely imperative for us to change our stance in Africa of being scared of AI to instead we must be the builders of AI. 

So that’s the first point. And that leads to the second point, which is that to be builders of AI you need very very fast computers to build them which people call compute.  They’re just referring to fast computers and typically fast computers need Nvidia GPUs right 80% if not more of AI computers in the world today are powered by Nvidia GPUs. 

And so those GPUs are not present in Africa today, extremely low, and we need to do something big to change that. Either it’s figuring out ways to improve the supply chain and be able to sell these high-end laptops in Africa, or it’s in the cloud service providers who are in Africa to buy more of these GPUs beyond the continent and then provide it to innovators and researchers on the continent at a more affordable rate. 

And now, as the co-founder of the Alliance for Africa’s Intelligence, which aims to drive African innovation on AI, and you’ve talked about the bias and discrimination caused by the algorithms from large language models, could you please discuss some of the projects that are currently under the Alliance for Africa’s Intelligence and what can be done to ensure that AI does not continue entrenching bias and discrimination among blacks and African perspectives in general

Yes, definitely. The simple answer is that we need to be builders. It’s not complicated at all. We have had so many complicated conversations about this for the past four or five years, and we seem to be missing the point. The whole point is we need to be builders. When we build the solutions, we reduce the bias question more than 50% because we will be supplying our own data ourselves. 

We will be interpreting the data in much better ways than anyone else can interpret the data. So that solves more than 50% of the problem. The remaining 50% of the problem is that we ourselves have our own biases. So we have to be aware of our own biases. In our own cultures, there are ways we’ve been biased in the past. The tendency of excluding women from work, we do it too in Africa. 

So when we build an AI HR tool in Africa, we should be aware that we don’t want it to replicate the same bias. And when you are aware of it, you can tweak the AI to behave differently from the way life used to be. 

It doesn’t have to be the exact way the data, the history was. You can design it to be different. Similarly, we also have some biases that relate to nationality or the tribe you’re from. We should be cognizant of these things when we build AI for Africa. 

But the answer is it’s not that we need to argue for policy and things like that. It’s that we need to be building AI tools. And to build AI tools, you need to train people. You need to provide fast compute, and then support them in terms of financing and all of that. 

And when we now come back to Alliance for AI and the ways we’re trying to support all these things happening, the first big thing we did back in 2020 was to create the first list of 100 AI startups in Africa and put them on a map. 

That was able to change people’s mindsets to think, maybe Africa is part of the AI race if there are 100 companies that are working in AI or using AI in Africa. So that was incredible. The second thing we did was to work with governments on AI policy, the same policies I’m talking about. 

To support that, well, if I took a step back, we put out a call in the United States for as many Africans who are already doing really incredible work in AI at universities, at companies, and from the black group, so literally a community, a WhatsApp community of really, really top AI people from Africa. 

We turned around and started talking to African governments and saying, you should be thinking about artificial intelligence. And whenever we get an engagement, we will involve Africans from that country in the engagement. So we ended up working with South Africa. We worked with Rwanda. We worked a little bit with Tunisia. We worked with Nigeria on their AI strategy, on their AI policy. And we worked with the Africa Union. 

We were able to connect them with their own people who are in the diaspora so that they can inform these decisions being made by our governments and that was really tremendous. The third thing that we decided to do was in the education space. We recognized that it was very difficult, it appeared to be very difficult for the Ministries of Education to make teaching AI a necessary thing for schools. They say it’s expensive to train teachers, it’s expensive to buy new textbooks and so many things. 

So we created an AI club program with our own curriculum and our own process for students to teach themselves in peer groups. And this way they have chapters and they’ve been teaching themselves about AI for many years now. So we have these chapters in Kenya, Tanzania, Rwanda, Tunisia and Ghana. And people have graduated and they have startups, they’re working with African companies, they’re working with  big global companies like Microsoft, Oracle, Goldman Sachs, and they’re becoming, we call them future makers. 

They are the ones who are changing things because they’ve learned now they can change their university and make them teach AI. And then the final thing I’ll mention is those fast computers, the GPUs, the compute. That’s the biggest thing that can happen in Africa today is the funding of GPUs, of fast supercomputers so that all the 

Many, like over 50,000 people have been trained in Africa to build AI now, but they don’t have fast computers to build. So it’s almost like you train someone to drive a car, but there’s no car to drive. The person has to leave the country. They have to leave the continent because there’s no car to drive. And so we need to bring the car to them. We need to bring fast supercomputers to them. And we hope that that project starts to have legs this year. 

On the same point of building our own AI to solve our own problems, you are the co-founder of Aura AI and you also support the ZINDI project. You’ve been seeing the work they’re doing over the years. What would you say has been the impact on data science in Africa and of solving problems from within using our own algorithms. How’s the journey been? 

Yes, there’s been a massive impact. Zindi.Africa is one of the most important AI organizations on the continent. It’s one that really blows people’s minds away. When you think of AI in Africa, you think, they’re just getting started. There’s nothing happening there. So the same way people’s minds were blown when they saw the Alliance for AI chart of the first 100 AI companies, that’s the same way they are blown when I ask them “How many data scientists do you think are on the Zindi platform?” The answers are like 100, 500. Well, no, it’s over 75,000 data scientists on the Zindi platform. 

It’s like a competition platform where companies that don’t have AI-capable people in their employee base or their staff, which is most companies, can come and host a competition on Zindi. And then the Zindians will compete for one week, two weeks, three weeks to solve the problem. And then you know after that time the company has a solution to their problems.

 Zindi provides a prize for winning this and that prize is typically cheaper than paying full time staff to work on a problem. And then secondly they can decide if they want to hire the team that builds the solutions. So this is a truly phenomenal solution that’s providing solutions for the Africa ecosystem. And those solutions have been transformational as organizations that host competitions continue to host more competitions because they are seeing the power of AI, the power of using your data to make decisions. 

You have organizations like banks. For instance, ABSA in South Africa brought their data over and the scientists were able to understand the data and figure out which kind of people are likely to commit fraud when they are collecting loans from the bank. And because of that, the bank was able to improve its decision making on which kind of people to perhaps not give loans to and which kind of people to give more loans to. 

It increased the financial inclusion because they were able to use this data. Another example is in the agriculture space where there was an organization that had pictures of many farms from drones. The scientists were able to look at those pictures and start to determine which pictures have plants that have some sort of diseases or pests. 

That way it can automate the detection of these things a lot earlier and then make proposals to farmers in small farms and big farms on what kind of solutions they need, whether it’s pesticides or different kinds of things that we need to use on the farm. These are just simple examples of some of the kind of projects that they’ve run on Zindi. 

It’s a demonstration of what can happen when we do this at scale, when we’re able to provide the kind of funding the scientists need, when we’re able to provide the kind of fast supercomputers these scientists need. 

One issue has been regulation of AI in Africa. We are only seeing very few countries developing AI strategies to prepare for AI take-off. My question is, what needs to be done by African governments so that they are at par with the rest of the world in terms of AI development in a safe mechanism while preserving ethics and data privacy?

The governments just need to focus on these exact things. The three things are data, talent, and compute. The government needs to figure out ways to invest in talent development. There are many, many NGOs and communities now working on talent development in AI for Africa, but they cannot reach the kind of scale that the government can reach. 

The government can support the Ministry of Education to formalize  the whole thing so that people are being taught in secondary schools and universities about AI. The government can also work on projects that involve data. 

People say Africa doesn’t have data. It’s not true. Africa has a lot of data. The challenge is the data is in silos. Different government organizations have their own data in different places. Different private organizations, hospitals have their own data. If there are incentives created so that data can be connected, they will now have an aggregate kind of data connected and in uniform. 

African governments should be the forcing function that drives and creates incentives so the ecosystem will do this for themselves. The governments need to invest in compute. The governments in Africa have done a lot of policy and strategy work so far, but none of them is investing their budgets in computers. 

I can’t say none. There are a few that have invested a little bit. They need to invest a lot more. They have a budget. They should use it for this. They should invest in data centres because if they don’t invest in compute, it’s all talk once again. It’s useless. 

They all say they are focusing 8 % on talent training. What are you training people for? There’s no car to drive. Buy them the car so they can do something. That’s where the government should focus on. 

You don’t focus on ethics and my ethics by talking. You focus on it by enabling your people to build. Right. So they should shift the focus away from talking about ethics and enabling their people to build. That’s the way to do it. 

Great. And over the years you have been in the AI industry, what do you think is the role of international collaborations in advancing AI research and applications in Africa? 

Collaborations will help. The role where we leverage them is to try and convince our private sector and our public sector, our governments to do what they’re supposed to do. When they do what they’re supposed to do, the need for international collaborations would be less. Their job is to support us. But right now we’re treating it as if their job is they’re the ones to do everything. That’s not correct. We need to step up and do our own parts. 

Our governments need to step up and do their own part. So as it is right now, the international donor community are the ones funding many things. They are funding talent training and data gathering. Little by little they are funding some provision of compute. So that’s what’s happening. And it’s good. We thank them and they should continue to do this. But we need the support of our own governments to do similar things. It’s our own responsibility to do it. 

We are seeing AI startups sprout from all corners of Africa right now, but they are struggling with attracting venture capital. What can be done to ensure that they get the same level of financing we see in Silicon Valley?

If the enabling environment is created by the government, these startups will have a better chance of building good solutions. And if you have a better chance of building good solutions, venture capital will run to you.

If venture capitalists see that there’s more talent to hire, more data to use, more fast support computers that are provided in an accessible and affordable way, they’ll know you can build AI solutions. So they will invest in you. But if they don’t see those three things, it’s like a waste of their money. Why are they giving you the money to invest? 

You know, when they give you the money, you struggle, you can’t find people to hire, you can’t find data to use, your compute access is extremely expensive. There was a study that was done last year that showed that it costs 30 times more to hire a data scientist in Kenya than in Germany. So it’s better to invest in a startup in Germany. So this is a very big problem and African governments need to do something about it. And yeah, that’s simply it. 

What is your ideal world run by AI?

Think of it less as a world run by AI and more of AI is just automation. AI is automating things that we already know how to do. And as humans running things, when you now know how to automate it, you do it faster using AI. For example, even in the past, a calculator is an example of something that’s similar to AI. 

Before calculators, you had to use your brain to do mathematics. You do one plus one equals two. You do big multiplication numbers in your head or you use a paper. But when you have a calculator, now you type it in and it will give you the answer.  That’s the same way with AI. 

If you have very, very big questions that you need to ask so you can do your work, you can now ask ChatGPT for the answer, for it to teach you how to think about the answer. And then you get the answer, right? And you get it in two minutes. But for someone else who doesn’t have ChatGPT, it would take them two weeks or two months to get the same answer. 

They are so much slower, you are so much faster. That’s how AI changes the whole thing. Another example for farmers, without AI as a small farmer, you have to go outside and dig and clear that whole place and remove weeds and things like that, or planting, planting the seeds. But with AI you can design an autonomous planting agent that you just press the button to go and plant everything in the whole farm for you. 

And by the way, this is not fiction. This is an actual company we invested in in Ghana that is doing automatic planting of seeds around the farm. That is what this is about is AI to automate things that we do as human beings. It’s  Africa’s chance of improving our economic situation. This is a very big chance for Africa and we should take it. 

AI is a great equalizer. It gives us tools that we can use to improve our economic situation. That’s what AI is and it’s you know, AI taking over the world.  But that’s not what is happening here in Africa and we have a chance to do things differently.

In this process of automating human tasks, we have seen several cases of AI hallucinations and we’ve also seen the potential for AI weaponization. How can the world control this? 

AI is like it’s your employee. It’s like your assistant. It’s helping you out. When you have an assistant or an intern, do you ever take the thing they give you without checking it? You’re not supposed to. You are the boss. You ask your employee or your intern to do something for you. When it brings it in, you give it feedback. It will change it. You give it feedback. It will change it. Eventually you like what is happening or you add your own final touch. That’s how it works. So with AI, it’s the same thing. 

That’s how it’s supposed to be the case. You change, you correct it when you see an issue. The point is not that AI is not human, it can’t do everything 100%. The point is AI gets you from 0 % to 80%. Saves you weeks, saves you so much money. Now your job is to complete it and complete the task. 

That’s how things work. So we should take the responsibility and we should be the boss, we’re the human being and AI is the intern. So let it do its own part and then you do your own part. That’s how you use AI. And the truth of the matter is every single week, the AI models get better and they make fewer errors because they learn. When people correct them, they learn over time, they make fewer and fewer errors. 

Now, the last question. Where do you see Africa in 2030 with this impact of AI? How do you see AI improving industries in Africa in the next five years? 

It depends completely on what we do. When it comes to AI, governments need to put a hand into it from educating a lot of people to enabling and creating enabling environments like when you have to import heavy equipment for AI. It’s very expensive and the governments have put an input duty on top of it that is extremely high. There’s nothing worse than this and our governments really need to stop doing it. 

It’s only the government that can change that. The private sector cannot. The governments have to become aware of this horrendous behavior and do something about it. And then in terms of providing a fast compute so our people can build, there has to be that big investment to subsidize the access to that. 

That is, these things have to change. If they don’t change by 2030, we’re going to have an Africa where there’s going to be AI from the rest of the world coming into Africa. We are going to continue to be consumers and users of AI. The AI we’re importing will harm people. The AI we’re importing would take away jobs because they will be much faster than the alternatives that are being built by our own people. 

When our own people’s products are not good enough, well, what’s the point? You fire everybody, right? You close down the company. We will lose so many jobs. And this is at the same time when our continent is going to have an explosion of population. We will have the highest population of youth in the world, all with no jobs. That’s going to be a very, terrible future. 

Our governments have to change. Our private sectors have to change and do the opposite, which starts by building these fast supercomputers, enabling the environment, and removing those import taxes. When you train enough people, our people will build AI solutions that solve African problems. These will create new markets that we are not even aware of right now. 

They will translate into millions of jobs. Even in the health space, I can imagine healthcare workers in villages with  smartphones that can answer many medical questions for patients in the villages. This would be a job for thousands of people. 

You will now have people in the creative industry that can use an AI tool to create better stories, create animation, and create movies. It reduces the cost of production by more than 100 times if you’re running the right application. You now have many more young people who can be more creative and create all these stories and animations of Africa. You’ll have so many jobs across so many industries because you enabled the African AI developers to build. 

That’s the future that I hope we get to by 2030. One where our young people who are of working age have jobs that are created using AI and we’re able to paint Africa in the way that we want the world to see it. I hope that we get there. 

Source : Impact Ai News

Send us your press releases to [email protected]