- 11/17/2023
Kate Kallot, founder and CEO of Nairobi-based startup Amini; Fatima Tambajang, NVIDIA’s head of developer relations, startups, and venture capital ecosystem for Africa and the Middle East; and moderator Eugene Anangwe, founder of East African Media Group, delved into AI, funding, and more during an “Innovation for Impact” panel at the inaugural TIME100 Africa Summit in Kigali, Rwanda, on Friday.
Category
🗞
NewsTranscript
00:00 Thank you for making time for this important conversation.
00:02 I wanted to start off with you, Kate.
00:07 You have received sustained global acclaim in tech
00:10 for social impact.
00:12 And I'm not just flattering here.
00:14 I think your work has been very notable
00:17 in advancing technology access across Africa.
00:21 I wanna understand the aspect of the access to tools,
00:25 resources, mentorship, which remain a big challenge
00:29 for those within the space that you operate in.
00:31 What is the missing link?
00:33 What has been the missing link in availing that?
00:36 Because it's been a challenge for decades.
00:39 We keep hearing those as major challenges
00:41 for people within the tech space.
00:43 What has been the missing link in closing that gap?
00:46 So I think the missing link has been big technology
00:49 actually caring for African developers.
00:52 There is one thing to note when you think about
00:54 how innovation is being driven across the world,
00:56 specifically AI innovation.
00:58 You look at the US, it's coming from big tech,
01:01 research lab, coming up with a new model,
01:03 a new compute infrastructure.
01:05 You look at China, very similar,
01:06 but then you look at Africa,
01:08 the innovation is driven a little bit differently.
01:10 There is an abundance of AI talent
01:13 and developers and startups here.
01:15 They just need the tools and the platform
01:18 to show what they're building.
01:20 And it's kind of an applied innovation.
01:22 They're using technology that's being built in the US
01:25 but applying it differently to our context
01:28 and localizing it for context.
01:30 So we need to make sure that we carve a platform
01:34 for these developers and the startups
01:36 to actually show the innovation
01:38 that they're driving to the world.
01:39 Interesting, I wanna hear from Fatima.
01:41 I mean, we are hearing that we have the developers
01:45 across the continent, but a lot of the developed solutions,
01:49 AI solutions often like we had from Kate,
01:54 are often imported or we get it from external developers.
01:58 How do you ensure that African developers
02:01 are at the heart of the party
02:04 and they're not actually just spectators
02:07 who get it from wherever it comes from?
02:09 How do you ensure that we have more developers
02:12 from within the African continent?
02:13 So I think as Kate mentioned, right,
02:15 I think on the continent,
02:16 a lot of things happens grassroots
02:18 and you have several communities.
02:20 Fun fact, Kate hired me at NVIDIA, right?
02:22 So at NVIDIA, they call me a baby Kate.
02:25 So she's had like a long history in big tech.
02:27 But you have Endavo, for example,
02:29 they have 32 chapters across the continent
02:32 in countries like Mali, Sudan, Chad,
02:36 South Africa and so forth.
02:37 And earlier this year, they had a summit
02:39 and in the summit, they had 1,200 developers
02:42 all showcasing work that they're doing.
02:44 You have the open AI's on ground and video's on ground,
02:46 deep mind was on ground,
02:48 showcasing the work that they're doing.
02:49 So I do think that the work is being done.
02:51 There just is invisibility into it.
02:54 Invisibility into it.
02:55 But Kate, does it matter where it is developed?
02:58 It does.
02:59 Why should it matter that Africa becomes a center
03:02 or a key player in the development of these AI solutions?
03:05 I actually think Africa already is.
03:08 And the reason is because like Fadima mentioned,
03:11 the AI innovation here is driven
03:13 from a grassroots standpoint.
03:14 So you just have to change your lens.
03:17 If you change your lens and if you think,
03:19 okay, I'm gonna consider innovation,
03:22 not a new platform or a new large language model,
03:25 but I'm gonna consider innovation,
03:27 a developer who has built a smart camera
03:30 that can be deployed in the bush
03:32 and identify animals completely remotely
03:35 with no access to connectivity.
03:36 That is innovation for us.
03:39 It might not be innovation for the rest of the world,
03:41 but here on our grounds, it matters.
03:44 So I actually think that we already leading
03:46 or we already at the forefront of AI innovation,
03:49 we just need to like remove our biases
03:52 and change the way we're looking at AI innovation.
03:54 Right, removing our biases.
03:55 What will it take?
03:56 How do we do that, Kate?
03:58 How do we do that?
03:59 Reverse the curve.
04:01 I mean, today we were just talking about this backstage.
04:03 A lot of the investments in AI and in deep tech
04:07 still come from the same places.
04:09 US, Europe, that's where most of the VCs are
04:12 and that's where most of the capital lies into, right?
04:15 African VCs are still extremely risk adverse
04:17 because they don't understand the technology.
04:19 So they would say, well, I'm more used to invest
04:22 into FinTech or e-commerce,
04:24 so this is what I feel comfortable with
04:25 and that's what I'm gonna go through.
04:27 So to remove the biases,
04:28 we actually have to all get around the table,
04:31 put experts at the heart,
04:33 whether it's companies from NVIDIA,
04:35 whether it's our startups from the ecosystem
04:37 and educate everybody on what AI is
04:40 and how AI innovation in Africa is being driven.
04:43 Yeah, Fatina, I've heard a lot,
04:45 especially in that particular regard
04:47 of where the funding comes from.
04:49 And I heard from one of the forums
04:52 that I've also been privileged to moderate
04:55 where the truths have been told
04:56 that the VC funding, external FDIs,
05:00 will dry up at some point.
05:02 And so we've already even started seeing that happening
05:05 in terms of the flow of capital
05:07 from the external people who support in that case.
05:11 How does the ecosystem in Africa prepare for this?
05:13 - So I think at NVIDIA what we do is,
05:15 well, we have a program called Inception, right?
05:17 And we have like 15,000 startups part of it.
05:19 Hugging Face, OpenAI, they're all part of it,
05:22 including Emini as well.
05:24 And what we do is we focus on technical capabilities
05:27 'cause that's what we find wins the race at the end.
05:30 And I think when people talk about AI in Africa,
05:33 there's a misconception that FinTech
05:35 is probably one of the biggest acquisitions that we've seen
05:37 and the case is it's InstaDeep, right?
05:39 So they were acquired earlier for 700 million.
05:41 And then the other question is, well, is it a one-off?
05:44 We have 320 startups at NVIDIA that we look after
05:48 when it comes to Africa.
05:49 And when it comes to VCs, the questions typically are,
05:52 what's the technology that they're using
05:53 and who are they building for?
05:55 And we find that from an early stage,
05:57 they're building for the global market.
05:58 So they're building in automated industries, agriculture,
06:03 and those are things that you can replicate across regions.
06:06 - Right, and if I could come back to you, Kate,
06:09 especially on the funding aspect,
06:11 but for women-led ventures,
06:14 what I know is 2022 female-led startups
06:17 raised 25 times less
06:20 of what their male-led counterparts did raise.
06:24 I wanna understand from you,
06:25 just give us an inside feel or look of how difficult
06:30 or how easy it is to raise funds as a female-led venture.
06:35 - The first round was extremely difficult.
06:38 I'm a woman, an African woman, building from Kenya
06:41 and doing deep tech and AI.
06:43 You've not heard that,
06:44 you don't hear that every day, right?
06:47 So the first one was extremely difficult for us
06:48 because originally we wanted to raise
06:51 from African investors
06:52 'cause we had that thesis as a team
06:54 that every time an African startup was successful,
06:57 the capital was coming from outside,
06:59 so the wealth was going outside.
07:01 So we really wanted to focus on African investors.
07:03 And what we found is that most of them
07:06 actually do not understand AI, deep tech,
07:09 and it was really hard.
07:10 So we had to do what we didn't wanna do
07:12 and go to investors who could understand the technology,
07:15 didn't really understand the African market,
07:17 but understood that there was a market outside of Africa
07:21 for what we were building.
07:22 So we were building a global company from Africa.
07:25 And then things started to ease up.
07:27 Our second round that we just closed
07:29 was way easier than this.
07:31 Thank you, Ty, because you gave us the platform
07:33 and the recognition.
07:35 And that helped us a lot
07:36 because all of the sudden you're shining the light
07:39 on a team that's purely African,
07:41 building from the continent,
07:43 but who has global customers outside,
07:45 but is also driving innovation from the continent
07:47 that can be reverse engineered somewhere else in the world.
07:50 And that really helped.
07:51 So it's all about the platform and the visibility
07:54 and the voice you give to African builders.
07:56 Right, and we wanna see more of Fatima's and more of Kate's.
08:00 And so I wanna hear from Fatima,
08:02 from where you sit, what's been the biggest challenge
08:05 and probably if you could just give us an experience
08:07 of trying to raise capital as a lady or as a woman
08:11 who is within the tech space
08:13 that often considered a male dominated space.
08:16 What's been your biggest challenge?
08:17 So I don't necessarily work within the sections
08:21 of raising capital, right?
08:22 So we partner with VCs
08:23 and then we connect the VCs to our startups.
08:26 And I think the biggest challenge is,
08:28 founders typically being asked,
08:29 are you sure you're building this from the continent?
08:31 The bias is an issue.
08:33 The other thing that's an issue
08:34 is there aren't many of us in this space.
08:36 Like I said, I randomly met Kate in New York
08:39 and we were speaking and she told me what she was doing.
08:42 I'm like, it's crazy, right?
08:43 And we're both born and raised in Europe
08:45 and then we both remitted ourselves back.
08:47 So I think that's probably the challenge, right?
08:49 Finding examples of what's possible.
08:53 The story of InstaDeep, it's not loud enough.
08:55 I don't think people know it enough.
08:56 I was in Nigeria last month and I was asking developers,
08:59 why aren't they building an AI?
09:00 And they're like, they don't have any examples.
09:02 But when it comes to FinTech,
09:03 they've been shown the possibilities, right?
09:05 Many people also don't know that Google has AI lab
09:08 in Accra, the first one.
09:10 It's been there since 2018.
09:13 Arm has five AI labs in South Africa.
09:15 They have five AI labs in Ghana.
09:17 So again, I think the challenge really is
09:19 creating visibility regarding all the work
09:22 that's being done in this space.
09:24 - Right.
09:25 Kate, I know we are built within the mentality
09:28 of seeing is believing.
09:30 And if we do not see, it's hard for us to believe.
09:32 And I wanted to just share with us
09:34 some of the practical examples of how AI,
09:38 let's not talk about the future,
09:39 let's just talk about now, let's be in the present,
09:41 of how it's already transforming or could transform
09:44 socioeconomic activities across the continent.
09:47 - So you know there are three things that,
09:48 as a continent, we have.
09:50 On one side, we have abundance of natural capital.
09:53 On the other side, we have abundance of young talent.
09:56 And then now, we have access to AI and data.
09:59 We'll talk about data as well
10:01 because there's no AI without data.
10:03 So the convergence of those three things
10:06 actually make it so that a lot of the applications
10:10 that are already coming from our grounds
10:12 are solving for socioeconomic issues,
10:15 environmental, access to financing.
10:20 And for me, my personal thesis is that AI, to be useful,
10:24 it has to be invisible.
10:26 So tech, for example, using data and AI
10:29 to make better or more informed decision making.
10:32 That decision making happens across the value chain.
10:35 That could be useful to a farmer
10:37 to understand better what's happening on its farm.
10:39 That could be useful to an insurance,
10:41 insurance works at the basis risk,
10:43 to calculate the basis risk
10:44 and provide agricultural insurance to our people.
10:47 That is useful to a bank to calculate a credit score
10:50 and give access to loans and microfinancing,
10:53 all the way to a government to also better food security
10:56 and understand the effect of climate change
10:58 on their country and a global private sector company.
11:01 Because once you have data and AI, you have transparency
11:04 and transparency creates trust
11:06 and trust accelerates economic investments.
11:10 So once you have, AI is actually applicable
11:13 throughout the entire value chain
11:14 and is applied today throughout the entire value chain.
11:17 It's just not as visible as it should be.
11:20 - Right, I wanna put the government institutions
11:24 off the hook for now, but I'll come back to them.
11:26 But Fatima, I wanna hear from you.
11:28 We've often asked what we want as developers
11:32 from government institutions, from policymakers,
11:35 from financiers, but I wanna twist that question and ask,
11:38 what is it that you think you'd want from developers?
11:42 - It's a great question.
11:44 I've never thought about it from that perspective.
11:47 I think it's just, you know, to understand
11:50 that when it comes to AI, it's delayed gratification, right?
11:53 So there's a saying that if you spend 10 hours
11:55 building an app, you're gonna get an app.
11:57 If you spend 10 hours tinkering, building an AI model,
12:01 you have no idea what you're going to get.
12:02 And that's also very difficult because the reality is
12:05 we are in a society where economic challenges are rife.
12:10 So you don't necessarily have the luxury
12:12 to like sit in a lab for three to five years
12:15 to tinker, to see what comes out of it.
12:16 But that's typically some of the challenges
12:19 that it comes when it comes to developers.
12:20 They feel like it takes too long.
12:23 The learning curve is extremely steep.
12:25 There aren't that many universities
12:26 that are offering comprehensive AI courses.
12:29 If they are, they don't have access to lab.
12:31 So they have to go through a lot of loopholes
12:35 when compared to building a FinTech app.
12:37 So it's just like trying to understand that, you know,
12:39 it really has a heavy research and development background
12:43 when you're choosing this field and the money will come,
12:46 but it will just come later on.
12:48 But it's, so I think that's typically
12:50 what we ask of developers.
12:51 - Right, Kate, what would you ask from government,
12:53 policy makers, just in a nutshell?
12:55 If they are to listen to you today,
12:56 if there's that one thing they need to change
12:58 to make the environment better for developers?
13:01 - I would say please do not copy and paste AI regulations
13:04 that are being done in the US,
13:05 because today that's the model they're following.
13:08 A few days ago, actually, Kenya released
13:11 the draft of an AI and robotics bill,
13:14 and the entire AI ecosystem in Kenya said,
13:18 "We cannot accept that."
13:19 So they are now trying to tax any innovation
13:23 that's coming from developers and startup
13:25 because they don't understand.
13:26 And what they're doing is taking a template
13:28 that's being created in the US and EU
13:31 with regards to data and AI regulations
13:34 and applying this on our continent,
13:36 not really realizing that, again,
13:38 the AI innovation on our continent is driven differently.
13:42 As much as you build an AI model in the US,
13:44 you apply it in Africa, it doesn't work
13:46 because it has to be localized,
13:48 regulations have to follow the same path.
13:50 - Right.
13:51 I don't want to turn you into a prophet,
13:53 but allow me to ask you, what do you see
13:55 in the next 10, five years?
13:57 How does the future look like in under 30 seconds?
13:59 - I can tell you what it looks like now,
14:00 healthcare and agri-tech.
14:02 So we're seeing AI solving sickle cell,
14:05 triple breast cancer, which affects black and brown people,
14:08 agri-tech, crop detection.
14:09 So I think that's what we're seeing
14:11 and we'll continue to see that in the future.
14:13 - How about you, Kate?
14:14 - So historically, we've been a geography
14:16 that has been consumer of technology.
14:18 So something is created there and then we use it,
14:21 we apply it.
14:22 Next 10 years, I hope that we're gonna flip the coin
14:25 and you will see more and more deep tech startup
14:28 like InstaDip really building infrastructure,
14:32 core technology infrastructure from the ground in Africa.
14:36 And for me, that's the hope.
14:37 - Right.
14:38 (audience applauding)
14:40 Thank you very much.
14:41 I think that's all the time we had for this conversation.
14:43 - Thank you.
14:44 - All right, thank you very much.
14:45 We can...
Recommended
8:56