Karim Beguir begins his interview with Rest of the world When bringing up Star WarsThis is something he does often, as he is from the remote Tunisian town of Tataouine, which lent its name and otherworldly desert landscapes to the film franchise.
For Beguir, this curiosity highlights how distant his hometown is from global technological hubs. It also evokes the image of an exciting and still untapped frontier.
Beguir, a mathematician, began his career in finance, working for banks in Europe and the United States. But in 2014, he left a comfortable job in London and returned to Tataouine. He wanted to do something more meaningful.
The most ambitious thing he could come up with was to create an AI company in his home country, where there were few precedents for tech entrepreneurship. He founded InstaDeep together with Zohra Slim, a Tunisian software engineer. Ten years later, the company has grown to more than 400 employees and offices in London, Paris, Berlin, Tunis, Lagos, Cape Town, Boston and San Francisco. InstaDeep is one of the few major AI startups based in Africa.
When the Covid-19 pandemic brought the world to a standstill, InstaDeep trained a large language model to accurately predict dangerous new variants before they spread. In 2023, InstaDeep was acquired for $682 million by BioNTech, a German pharmaceutical company that created the first Covid-19 vaccine approved by the US Food and Drug Administration. The deal was the largest technology acquisition ever made in Africa.
For Beguir, going into the health sector would be a kind of homecoming: he would follow in the footsteps of his father, a doctor.
In August, Beguir spoke with Rest of the world via video call From London, where he currently lives, he spoke about the beginnings of InstaDeep, how the pandemic boosted the company and why he believes it is vital to create artificial intelligence companies in Africa.
This interview has been edited for brevity and clarity.
Tell me about when you started InstaDeep in 2014. How did the local ecosystem receive the idea of a Tunisian AI startup?
We got into the world of AI after reading an article about convolutional neural networks, which is basically visual AI. My co-founder Zohra and I decided to try our luck and built a virtual AI system that would allow people to scan luxury items, such as watches and handbags, to determine whether they were original or not. A year later, we sold it to a client in London.
At the time, no one could believe that we were talking about developing artificial intelligence in Tunisia. There was almost no startup ecosystem in the country and creating an AI startup seemed like a difficult task.
Initially, we didn’t plan to expand internationally. We tried to sell our first products in Tunisia, but we realized that the local market was not ready for AI. For people, it was pure science fiction and traditional companies did not want to pay for our technology. They did not understand its value; we only managed to generate a few thousand dollars in sales locally. So we started looking abroad.
When did you realize you had created a company that would last?
When I left London in 2014, I gave myself five years to try and get started. I started to see a glimpse of what we have become when I was invited to meet Mark Zuckerberg in Silicon Valley in 2017. Then, when we raised a Series A round of funding led by AfricInvest, it became clear that this very interesting project had real potential.
At the time, we were building systems that optimized logistics in trucking and carpooling services. DeepMind had created an AI that could beat humans at complex games, and we found a way to apply their game concept to real-world problems. We were among the first in the world, if not the first, to apply DeepMind’s innovation to industrial optimization.
Our research papers attracted the attention of the DeepMind team, leading to collaborations with DeepMind, Google, and the Massachusetts Institute of Technology.
Let’s talk about COVID-19. Everything comes to a standstill and InstaDeep becomes a global sensation after creating a system that predicts which COVID-19 variants could pose a health risk. Can you walk me through that moment?
InstaDeep’s competitive advantage is that we are not focused on one sector.
When COVID-19 hit, we were already working with BioNTech on personalized cancer vaccines. Very soon, the BioNTech team had a COVID-19 vaccine in place. Then, we started seeing new variants of the virus appearing every day – at one point, the number of variants sequenced every week was over 10,000. And I was thinking, “Who can follow this in real time?”
It was clear to me that only an AI system could do this. So we built and trained an AI system that was essentially a generative model. We took transformers that had already been pre-trained on all the existing proteins and then altered the language of SARS-CoV-2. We tried to see if we could predict with some level of accuracy whether or not something might be high risk. It turned out that it worked. All of the variants that we identified as potentially dangerous were subsequently confirmed as being of concern by the World Health Organization.
Was that the beginning of the acquisition conversation with BioNTech?
It was a very important moment, because when we started working with them, it was a one-time project that would last a certain number of months. They were testing us. The project was successful, so we expanded to do more projects. The more time passed, the more we demonstrated the quality of what we were capable of doing and showed them our creativity, which is extremely important in AI. We created a great collaboration with our colleagues at BioNTech and the decision to join forces was a natural one.
-
What’s the first thing you do in the morning?
- Read the latest news on AI in X
-
What is your favorite Netflix show?
- The Office
-
What was the last song you listened to?
- Kavinsky’s Night Call
-
What is the food you miss the most when you are away from home?
- Mloukhia
-
If you had to start InstaDeep from scratch in another country, where would it be?
- If it wasn’t Tunisia, I don’t think I would ever do it.
-
How do you relax?
- History of reading
Lightning round
The mainstream conception of AI is generative AI, such as ChatGPT, Midjourney or Gemini. These are all programs that have demonstrated a bias towards developing countries. How is InstaDeep different?
The key ingredient is to make sure that those powerful technologies are actually adopted, used and taken up by different communities around the world. However, those technologies are developed in the United States or in Asia with their own biases. Africans, for example, are far from being the creators and are not the main target audience. That is why the feedback loop from here does not work.
At InstaDeep, we stay away from controversial applications of AI. We focus on non-controversial ideas with net benefits. The only visual AI system we developed was in the early days and was intended solely for objects, not people.
However, our approach to combating biases or imbalances in the AI world is to empower young people, who would otherwise not have the opportunity to develop their own projects for their own countries. We have offices in Tunis, Lagos, Kigali and Cape Town. Empowering young talents in Africa is about creating a tighter feedback loop and also ensuring that the incredible economic value that is going to be generated in the coming years is also leveraged to some extent in the country.
You are a big proponent of building in Africa, but in 2015 you moved InstaDeep’s headquarters from Tunisia to London. Why?
InstaDeep raised a $100 million Series B funding round in 2022. Because of how Tunisian law works for startups, I don’t think we would have been able to raise that amount locally. In this case, it was necessary to incorporate in the UK, where I had lived and knew the ecosystem. The UK ecosystem is extremely dynamic, but more importantly, it is easier to attract high-level customers and investors as you scale.
What do you think are the advantages of building from Africa?
There are two big advantages: one is the community and the second is the opportunity to innovate. The African machine learning community is very dynamic. For example, the Deep Learning Indaba, the annual gathering of African artificial intelligence researchers in different countries, was cited as a model for the world to adopt. Secondly, technology and digital are still in their infancy on the continent, and that leaves room for innovation. Today, in the United States, it is difficult to create something new.
#InstaDeep #Africas #successful #startup