Artificial intelligence: accompanying partner countries towards an inclusive and sustainable future

Artificial intelligence as a catalyst for the just transition

During this day of discussion, a number of speakers shared their vision of AI and the opportunities it offers for the future. Reina Otsuka, speaking on behalf of the United Nations Development Programme, gave examples of how AI can be used as a tool to tackle environmental crime, especially in coastal areas through the surveillance of coastlines and maritime traffic.

Darlington Akogo, Chief Executive Officer of minoHealth AI Labs in Ghana, addressed another crucial issue for the sustainable application of AI: “We help farmers better manage crops and control crop diseases caused by disruptions in seasons by using drones which analyse crops, and provide information about the health of corn, for example. All these data are analysed and make it possible to achieve better yields and more sustainable growth, especially for small producers.”  

Global ambitions for AI: gaps, expectations and fostering better cooperation

The development of AI constitutes a new approach to cooperation between countries of the Global South and with traditional partners, to ensure that it can be used and that there are lasting benefits, in particular in Africa. In this respect, Aissatou Jane Ndiaye, Director of ICTs at the Ministry of Communication, Telecommunications and Digital Economy of Senegal, calls for more people-centric public policies to tackle unemployment and digital illiteracy. By coordinating action at the national and local level, Senegal is in this regard at the forefront of the issue of innovation/research by supporting an essential spirit of entrepreneurship.

There are two challenges: ensure that AI is ethical and promote cooperation models to take the lead in these new fields of technology.

This is also called for by Lacina Koné, Director General and Chief Executive Officer of Smart Africa, a Rwandan multi-stakeholder platform. He believes that AI is an excellent opportunity for creation in Africa. Firstly, for job creation, as it is not about destroying jobs but creating them.

In health, education and all the sectors of the economy, it concerns 230 million jobs for young people by 2030 on a continent where 1.4 billion people are under the age of 30 (Sub-Saharan Africa).
“Africa needs a harmonisation of policies at all levels: NGOs, civil society and governments. There’s a need to review the manner in which digital transformations are implemented within the very functioning of governmental mechanisms, and ensure that there is a better inclusion of each national specificity. There also needs to be continuity with the private sector to ensure that AI is sustainable. It’s a priority for the continent to motivate young people to take this route, and the root of this lies in AI investment in education to get them interested in these subjects and, above all, give them the means to get better training,” he said.

Take account of the wealth of languages and cultures in this digital challenge

One of the major subjects for the future deployment of AI and the integration of these technologies into development policies is the issue of the use of languages, and the language barrier that all the actors involved in the deployment of AI need to take into account.
This issue of AI for all, which was a recurrent theme on 26 March, comes down to countering the digital divide by taking all languages into account. This issue is raised with the translation tools for rare languages, or languages that are far removed from English.

Assaitou Jane Ndiaye drew attention to the difficulty of mobilising farmers (only one example among others) whose language is neither English or French.

Lacina Koné believes that this is an important issue, in particular for the development programmes implemented on this matter. In Africa, more than 350 million people express themselves in local languages, on the fringes of the languages used by the current AI programmes. This raises the question of how to give them a greater voice in the future. 

Sid Ravinutala, IDinsight, illustrated this point several times with the leading example of India which has managed to develop more inclusive models through the study of languages in partnership with a group of universities: “An organisation in India collects linguistic corpora through recordings, and this highlights the wealth of languages and democratises them in GenAI. This has resulted in a pan-Indian model which works in all languages thanks to the joint collaboration with partner universities. This model can be replicated in Nigeria, for example.”

But an extension of AI to all languages is not necessarily feasible and advisable depending on the type of application. François Yvon, Director of Research at CNRS, used an example given by Sid Ravinutala, on the issue of using models in Swahili in the context of maternal health policies in Kenya, to comment on the relevance of a widespread deployment: “This works well in the present case, but in fact, in some cases Swahili isn’t taken into account despite the audience, and it involves deploying substantial amounts of financing. We need to ask ourselves about the funding priorities for certain countries and certain enterprises. Is school not simply the highest priority?” 

Innovation through AI: a driver for the public and private sectors

Samar Sobeih, a member of the OECD AI Policy Observatory and Partnerships Manager at the Egyptian Ministry of Communications and Information Technology, commented on the complementarity between the private sector and the public sector: “Public-private partnerships drive innovation forward by capitalising on the issues and interests of both sectors, they are complementary. The private sector offers financing and its expertise, and the public sector gives scope on the regulatory side and contributes to the development of partnerships at national level. We need to bring together start-ups, universities and research to move towards a more ethical and more stable AI.”

Agnès Leersnyder, Chief Executive Officer of Future4Care, highlighted the importance of innovation through AI in the field of medicine. Future4Care, whose founders included Sanofi and CapGemini, started from the premise that the digital health market was sorely lacking innovation.
“There is still a problem and it’s the fear of innovation, and especially of the use of data. This complicates the adoption of new technologies and innovations. It is crucial to call on experts to deal with the competition and cultivate this spirit of innovation. Today, we’re working with 50 start-ups and 7 new partners. They’re all working in and for AI. It reduces the cost of R&D and improves access to appropriate medication, for example, by reproducing the effects of a molecule on a particular body according to its characteristics. It’s fascinating,” she says.

Paul Duan, co-founder of Bayes Impact, believes that this issue of innovation is essential, but that it needs an appropriate framework. “Innovation is a good thing, but once businesses succeed, we need to be able to permanently implement the functioning of these new systems. We also need to draw on our passion, be open, and work collaboratively to make sure that these innovations don’t come up against exclusive preserves in certain sectors, such as the food industry.”  

This passion and commitment is also strong in universities, as pointed out by Lawrence Nderu, Head of the Department of Computing at Jomo Kenyatta University: “We need to support AI students with their abundance of ideas to set up companies and make sure they are sustainable. We need to combine this with partnerships, a successful incubation, and the necessary support.”

To wrap up this high-level event, Lacina Koné, Director General and Chief Executive Officer of Smart Africa, expressed confidence in the future of AI and its deployment in all African countries, despite the many ethical and strategic challenges: “AI for sustainability and inclusion, with ambition and respecting regulations. 

But where are we going? This question must direct us, guide us, and encourage us to establish solid foundations. We really need to develop the concept of multilingual AI to close the current gaps and this digital divide. In Africa, as in Europe, a leapfrog is necessary to take back the position of leadership in new technologies, and this requires cooperation between the EU, the D4D Hub, and African partners…”

“We’re going to bring concrete input to these discussions by launching an Advisory Group with Expertise France, which will also get all the stakeholders on the continent to work for sustainable and ethical AI, because if you cross a river alone, you’ll be eaten by crocodiles,” he said in conclusion, adding a touch of humour, to initiate a continuation of the cooperation between Europe and Africa.