Generative AI Talk Series Session 2 - Unlocking the Future: Generative AI Reshaping Development Cooperation
September 29, 2023
“I think up until this point, the main drivers of AI technology and the main success stories are around commercialization and driving capital means and entertainment. And I can't wait until the point where it's really solving problems on the ground and really addressing issues of inequalities, especially within the majority world”
- Pelonomi Moiloa,
Founder, Lelapa AI
NB: We have included an audio-only version of the event for those who may have issues with video playback or just like to enjoy podcasts on the go - please find the audio file under the videos.
The breakneck speed of advances in Generative AI over the last two years has taken the world by storm: In a short while, it has become clear that Generative AI has the power to shape not just the future of work, but also of all levels of society, due to its wide-reaching, varied impact affecting everybody. GIZ, with its core competencies of international cooperation, development knowledge, skill, and management expertise, stands uniquely poised to incorporate global voices into the dialogue on Generative AI and contribute to its inclusive and responsible development and application.
The second session of the Generative AI Talk Series, hosted by the GIZ Data Lab and Data Service Center (DSC), convened to discuss these issues and more. The session, titled “Unlocking the Future: Generative AI Reshaping Development Cooperation”, followed on from the previous edition of the Generative AI Talk Series which had featured Professor Luciano Floridi discussing the ethics of Artificial Intelligence.
The panel featured Ms. Pelonomi Moiloa, CEO of Lelapa AI, an AI start-up for Africans, by Africans, solving African problems; Ms. Lane Dilg, Head of Global Government Partnerships at OpenAI, the research lab behind ChatGPT; and Ms. Frederike Kaltheuner, Expert on Tech Regulation and Author, formerly Director of Tech & Human Rights Division at Human Rights Watch and of European AI Fund. The panel was moderated by Ms. Catherine Vogel, Coordinator of the GIZ Data Lab.
The event began with GIZ Board Member Ms. Ingrid-Gabriela Hoven setting the scene for the conversation by detailing how the challenges of the concentration of computing power could exacerbate digital inequalities in an already unequal world but also emphasised the potential of utilising this novel technology in achieving the 2030 Sustainable Development Goals (SDGs). An integral point of her address was the need to adopt people-centred approaches to utilising Generative AI in the sphere of development cooperation, concomitant to the SDG principle of ensuring that no one is left behind.
The panel, drawing from their deep sectoral savvy, provided an amalgamation of opinions on how they have viewed the last 9 months of Generative AI, and their hopes for the future of the technology. Ms. Moiloa and Ms. Kaltheuner expressed a mixture of multifaceted hope for the future with a slight mix of preoccupation. Both speakers stressed the importance of addressing the power imbalance that developing countries currently face in Generative AI. An equal AI system is one that is people-centred, and one which needs to work towards better including diverse points of view to reflect the collective intelligence of all humanity. Ms. Dilg, as a representative of a company that has been instrumental in shaping the Generative AI landscape as we currently know it, highlighted that OpenAI was founded as a Research & Development Laboratory first and foremost and their commitment to that remains paramount, as highlighted in the company’s mission and charter. She went on to outline that OpenAI is an entity focussed on learning, collaboration, and close dialogue with all its stakeholders.
When quizzed on the impact of Generative AI on the SDGs, the panel highlighted how education, particularly in local languages, presents perhaps the biggest enhancement to human capital that AI can bring. Both Ms. Moiloa and Ms. Dilg highlighted that bridging the digital divide through languages other than English will bring technological bases to individuals who would otherwise have been excluded from harnessing its powers. Ms. Kaltheuner stressed that a particular challenge to keep in mind when designing Generative AI to accomplish SDGs is that the technology can augment previous inequities and heighten exclusions. It is important to centre transparency and accountability in any system that augments its work with Generative AI, she said.
With the panel highlighting education and knowledge as crucial to enhancing human development when combined with AI, event moderator Catherine Vogel pushed the participants to hypothesize on what the way forward for a knowledge and expertise provider like GIZ would look like. The answer was that AI can enhance but not replace the work GIZ does in global cooperation. With its strengths in convening different stakeholders and stitching disparate pieces together in a cohesive whole, GIZ can look forward to an AI strategy whose bedrock is built on collaborating with communities and in building up local ecosystems - a point echoed in Ms. Hoven’s opening address.
With more than 900 participants, the event threw up a myriad of ideas to reflect, collaborate, and work towards. Ms. Hoven, while delivering closing remarks, mentioned that while she expected the event to leave her with more questions and points to ponder, it has also brought with it a sense of urgency to build AI systems that can balance power and give people a voice.