Wednesday morning’s plenary session included a debate, chaired by Professor Catherine Hankins, Amsterdam Institute for Global Health and Development (AIGHD), The Netherlands), in which two panellists argued in favour of urgent investment and two argued against (their positions did not necessarily reflect their actual views). Speaking in favour:
- Dr Annie Hartley (Laboratory for intelligent Global Health and Humanitarian Response Technologies (LiGHT), Switzerland)
- Darlington Akogo (minoHealth AI Labs, Ghana)
Speaking against:
- Adama Ibrahim (Novo Nordisk, Denmark)
- Dr Youssef Travaly (EDPU AFRICA, Rwanda)
Before the debate began, a Mentimeter poll of the audience found a large majority in favour of the motion.
All in on AI
Kicking things off, Annie highlighted that ‘urgency’ was the critical word. Millions of preventable deaths were occurring in Africa and AI could radically cut their numbers. She suggested that information was the most valuable commodity for physicians – specifically, “the right information at the right time and in the right place”. This is what AI could deliver.
Young doctors are already using AI, she noted, but in an unregulated way and without contextualisation for African settings. Africa needed to generate the data to create Africa-specific tools, leapfrogging legacy systems: “It is time for us not just to follow – Africa is in a good position to lead,” she concluded.
Not so fast
Adama countered by querying whether AI really was Africa’s top priority at this moment. What often holds back medicine and leads to avoidable deaths is the lack of basics – such as electricity, clean water, education, and a trained workforce. Should AI take precedence over these? “We need to build the road before we build the car,” she suggested.
She also highlighted a number of risks associated with AI. AI is fuelled by data but Africa has weak systems for protecting data; existing data have built-in biases; AI could drive mass employment; cybersecurity challenges need to be addressed; and AI can have massive environmental impacts.
From the floor
Audience members contributed to the debate, raising additional issues such as:
- The need to prepare professional to use AI.
- AI is already being used, but is not ‘owned in Africa’.
- Coordinated investment across the continent is needed to avoid a dependency on donors.
- Failure to invest could broaden the equity gap between North and South.
- What is the business model and what is the role of the private sector?
The final word
Responding to the first round of comments, Darlington noted that Africa needs 50 million more health workers to achieve universal health coverage. This is implausible but the challenge could be substantially addressed by increased productivity driven by IA use. The longer the delay, the more lives would be lost, he suggested. The private sector had a critical role to play, he added, providing impetus to Africa’s tech industry.
Youssef highlighted some key challenges, including the environmental impact of data centres, and practical data issues such as the lack of unique identifiers for people and, in many settings, specific addresses. She suggested that upstream investment was still needed to create a conducive environment for AI.
A further lightning exchange cited the risk of missing an opportunity, but also the point that Africa might be better placed not following the herd: “We need to do what’s right for us,” said Adama.
And the winner is…
A second Mentimeter poll revealed a dramatic turnaround. The ‘for’ vote nudged up slightly, but the ‘against’ vote leapt dramatically. By the rules of the competition, the ‘against’ team therefore won. As the ‘for’ team gained most votes, though, perhaps both teams were winners in the end…
The thought-provoking comments highlight a real dilemma. Not many seriously doubt that AI is going to have a dramatic impact on many aspects of life, including science and medicine, but exactly how to ride the AI wave is not yet clear.