Using AI to Bridge Rather than Widen Global Gaps: A Well-Wishing Utopia?

I am sitting with friends in the courtyard of a house in the capital city of a West African country. In the background, the (not so gentle) noise of a generator, running almost constantly as the country is grappling with an electricity crisis. Some of our friends are calling to say they will not be able to join us because of a shortage of petrol. Among those present, discussions revolve around the electricity and fuel shortages, and the fear of new terrorist attacks. AI and the debates surrounding it are conspicuously absent from that list of topics. 

​

A few days later I am back in Switzerland, listening to friends talking about their fears of AI making their jobs redundant, and expressing their worries that their children feel more comfortable talking about school bullying or suicidal thoughts to conversational AI tools than with them. Countries of the “Global South” and their realities are conspicuously absent from these discussions. At least on the surface. Because the lithium used in the batteries of the electric vehicles my friends have driven today in part comes from these areas of the world. Everything is related. So what role can and will we make AI play in bridging or widening this disconnect?

​

I have been growing increasingly frustrated with the absence of consideration of the “Global South” in discussions around the planning, development and implementation of “emerging technologies”. It seemed to me that these countries and their realities were ignored. However, I have learned from my misconceptions (and hopefully AI can do the same) and am trying to balance my somewhat subjective opinion. This started with the question of electric and self-driving vehicles. My initial reaction was: how could this even be workable in countries where so many households, and entire regions, are not electrified? And roads are unpaved/unmarked/unmapped? Well, whilst it is still rare to see electric vehicles in the streets of Bamako, Ouagadougou, Niamey (to cite just a few), we are seeing the arrival of electric motorcycles. Many houses now are equipped with solar panels. Projects using wind-power generators are also being developed. 

​

It is still slightly infuriating to know that lithium and other rare earths are being extracted from these countries, exported to other countries where they are transformed, then returned in their new forms (batteries, solar panels, etc.) at a sometimes prohibitive price…

​

But maybe AI could help with that? For example, by helping develop training programs for a skilled local workforce, making available the necessary technologies at a lower cost, etc.

​

Also, by helping local organizations and entities plan, develop and implement their own projects. This has really been frustrating me over the years (I admit I do get easily frustrated): the number of projects that are (often with the best of intentions) thought of and designed in European countries or the US, with foreign funding and insufficient knowledge or consideration of the local specificities and the real (as opposed to perceived) local needs. The latest example of this that I have witnessed in a West African country. The project was to gift/deliver a fleet of electric motorcycles to the municipality of a district of a West African capital city, to be used as “moto-taxis”. On the condition that these vehicles be driven exclusively by women. The project failed to start as they could not find enough women volunteers. An in-depth preliminary study or deeper knowledge of the local context would have identified that there are almost no women moto-taxis in the city. 

​

Whilst the abrupt stop of USAID and the cuttings in the budget for foreign aid globally over the last years have had very significant negative impacts on projects run in the Global South, at the same time this has caused a positive offshoot. Local organizations, faced with the disappearance of their usual partners and funding, have had to rethink themselves. Here again, the use of AI could be beneficial, making technologies, project design and implementation tools more available and affordable to local organizations.

​

AI could also be used to collect accurate data, thus “healing” itself from its current bias on countries in the Global South. Indeed, when one hears of these countries, it is, more often than not, because of an armed conflict, a coup, an epidemic outbreak, a famine, or an allegedly rigged presidential election. Today there is insufficient (in number and quality) data on these countries and their population. Very often important sectors (health, administration, education) do not have electronic records and still rely on pen and paper. 

​

This obviously prompts questions about the willingness of states to share data, the sovereignty of states over the data, the respect of individuals’ right to privacy and control over their personal data. But, wait, isn’t this also one of the main issues being debated in Europe and the US? And which could certainly benefit from a global discussion, taking all countries and their specificities into consideration.

 

I sincerely hope that in a few years’ time I can start another article with: “I am sitting with friends in a courtyard surrounded by no other sounds than the local music and the gentle humming of electric vehicles proudly made in that country by qualified local workers, using local natural resources and renewable energies”. And I would welcome AI tools’ help in achieving that.