Alessandro Crimi and Disruptive Innovation in Machine Learning from Africa

Yes, it is happening

Machine learning and data science is an interdisciplinary discipline combining several fields as statistics, computer science and often linguistics.

Many disruptive technologies we are seeing popping up every day are based on some machine learning solutions. There are many companies helping lives in the medical field, finance and even pushing further self-driving cars. Cloud based service technologies as AWS have been booming since the last decade. Little is known from initiatives from Africa, as it might sound counterintuitive.

We spoke with Dr. Alessandro Crimi, a scientist currently mainly resident in Poland at Sano center for computational medicine (https://bam.sano.science/) , but constantly working in Ghana, Rwanda and South-Africa, teaching, doing projects and pushing for startups.

SVT: Hi Alessandro, are there relevant startups in Africa in the field of machine learning and data science?

A: Obviously it is not Palo Alto, but thanks to large venture capitals in Sub-Sahara Africa and focus on startups from there, we have the first stars. For example DataProphet (https://dataprophet.com/) from CapeTown (South Africa) last year received a significant investment of an undisclosed amount from Yellowwoods Capital Holding. They are mainly focused on autonomous manufacturing, but also moving into AI based finance and insurance. Also from CapeTown there is Aerobotics, drone based farming. They got 27 millions USD as funding, not bad. From Nigeria there is Kudi.ai – now called Nomba- with messaging and safe transfer of money. Just to cite few.

SVT: What is your relationship with machine learning and Sub-saharan Africa

A: My love story with Ghana is actually quite old. In 2012, I went for the first time to the Ghanaian center of the African Institute for Mathematical Science (AIMS) to teach an introductory course on machine learning. At that time nobody, especially in Sub-Saharan countries, knew what machine learning was. The name sounded quirky to students, so I changed it into “pattern recognition”, but most often I had to explain what we were doing because also “pattern recognition” sounded weird. The first students were very interested but found it still a bit weird. Luckily, with time the idea of datascience reached the every corner of the globe, and then there was almost a queue to attend a course on machine learning. Then from Ghana, I moved doing projects and teaching in South Africa, Rwanda, Zimbabwe, etc.

SVT: What about other universities in Ghana and Sub-Saharan Africa?

A: At that time there wasn’t much. Now, in South Africa there are several programs and younger institutes are hiring young talents and there are plenty of possibilities. AIMS has also launched a course specialized in machine learning, : the African Masters of Machine Intelligence (AMMI) with  high profile lecturers as Yann LeCun and Joshua Bengio, Google is also involved. At research level, I bet many people are using machine learning there, not only companies.

SVT: Are there any connections between research and company to translate research into a profitable business?

A: If you mean incubators and accelerators, they are at every corner. There is an organization called Zindi (https://zindi.africa), they put a lot of challenges and hackathon on their website and the purpose is to connect talents and companies. I am very intrigued at the moment by Deep Learning Indaba (https://deeplearningindaba.com/2022/). It is first of all conference held on purpose in Africa, this year in Tunisia. They also propose mentorship programs to plan careers in machine learning. I am currently a mentor in this program. Indaba is also opening many local chapters to promote small conference and transition mentorship from research and companies they are called IndabaX.

SVT: Ok, it seems they are like the rest of the world contrarily to what people may think. When are they building self driving cars?

A: Building them? I am afraid it is too early, you need further critical mass in terms of talents and ventures, maybe they will catch up when we will transition to vertical take-off cars. But the market of users is there. I was reading a survey that most Ghanian citizens will love to buy a self-driving cars.

SVT: Other disruptive technologies.

A: I have seen a lot of interest in quantum computing. Again, to build a quantum computer in Africa from scratch might be too much, but the talents to use it are there, and with cloud computing you can have quantum expert in Lagos (Nigeria) and use the IBM quantum computers in Melbourne. There are many initiative like OneQuantum Africa (https://onequantum.org/africa/). I am involved into another project called Quantum Leap (https://quantumleapafrica.org/) in Rwanda. We have some PhD students working on projects related to health, data science and quantum computing.

SVT: What is the local mindset of success and making impact as machine learning entrepreneurs?

A: If I have to think first of all about the people I met, I would say they are nerds. The economy of some of these countries has been behind for centuries, and now people want to be like Elon Musk (who is African by the way). So, love for technology and innovation is the major strength I have seen, obviously making impact is probably there but people are relatively more humble than in Europe or US, so they don’t flaunt it as typical western entrepreneurs do. About success mindset, I don’t know. I am not African, I come from another culture. I notice stronger differences in other aspects like religiosity and sense of community, which are particularly strong in Sub-Sahara. Maybe the successful entrepreneur mindset is blended with those aspects.

SVT: Any final remarks?

A: Concluding, I think that Sub-Saharan Africa is still a young continent with plenty of opportunities even if they arise by social and health issues. To young talents I recommend to avoid piling up several degrees (something I have seen a lot), and be more aware of the local venture capitals you have in your country.