Expert Panel Reflects on Emerging Uses of Technology for Development
By Aditi Ramesh and Andrew J. Zahuranec
Five years since the United Nations General Assembly set forth the Sustainable Development Goals, people from all sectors of society have dedicated themselves to seeing them fulfilled. While significant achievements have come about in this time — with poverty in decline and increased access to drinking water — progress remains limited in many parts of the globe due to complex challenges, a lack of political will, and limited resources. Many have begun to look at how they might overcome these problems in part by leveraging new tools and technologies.
On Friday, November 13th, The GovLab and Agence Française de Développement (AFD; French Development Agency) presented one such examination. As part of the Finance in Common Summit, held alongside the 3rd Paris Peace Forum, The GovLab’s Stefaan Verhulst and AFD’s Peter Addo led a panel discussion on “Emerging Uses of Technology for Development: A New Intelligence Paradigm.” The event was informed by a new draft policy paper that seeks to provide a new paradigm for thinking about emerging uses of technology in development; describes the value and risks of applying new technologies; and details principles and considerations to consider when assessing projects.
Several distinguished experts and practitioners working at the interesting of development, innovation and technology, including Shachee Doshi (USAID), Marc Lepage (Asian Development Bank), Nii Quaynor (Ghana Dot Com), and Carlos Santiso (Latin American Development Bank) joined Stefaan and Peter to discuss ways this draft document could be improved prior to final revisions on November 25th. The video of the full panel as well as highlights of the conversation are below:
Introducing a New Intelligence Paradigm
The panel opened with a brief overview of the draft report by The GovLab’s Stefaan Verhulst. As he noted, “Emerging Uses of Technology for Development: A New Intelligence Paradigm” is the culmination of months of research into the ways technology can complement ongoing development work. Using real-world examples, the piece identifies four forms of intelligence enabled through technology that can address complex and dynamic challenges in low- and middle-income countries:
- Data intelligence, born from datafication, analytics, and advances in computer sciences, uses data analysis or visualization to support decision-making;
- Artificial intelligence, aiming at the automation of activities that are typically associated with human thinking, such as learning, decision-making, or natural language processing;
- Collective intelligence, the product of advances in communications technologies, social media and platform models, mines the wisdom of crowds to find ideas and insights beyond the traditional boundaries of public institutions and expertise; and
- Embodied intelligence, derived from breakthroughs in the design and computing power of embedded systems, enables these three intelligences to be deployed as physical agents capable of acting autonomously in the real world.
Each intelligence in this paradigm can be valuable for different purposes, whether that be enabling better assessments and experimentation (data intelligence), improving efficiency (artificial intelligence), enabling more legitimate decisions (collective intelligence), or facilitating the delivery of new services (embodied intelligence). However, as Stefaan discussed, the impact of these tools depends on their appropriateness, the context, and risk mitigation strategies.
Reflections on the Four Intelligences
For the panelists, this framework provided significant value because it codified emerging practices in the areas they worked. The four intelligences paradigm and the broader report helped to center the discussion on emerging technologies in communities.
“It’s a good way to focus attention on the purpose of applying technology [rather] than the technology itself,” said Nii Quaynor on the value of the framework. “The intelligences provide, perhaps, a pathway to achieving the SDGs by supporting decisions [and] helping [communities] to demand better outcomes.”
He went on to describe how he had seen the technologies described used in Ghana to address corruption and enhance transparency. He also spoke about the value of programmable money, an emerging innovation in finance not discussed in the report, which has implications for smart contracts and digital currencies.
Carlos Santiso and Marc Lepage echoed several of these remarks. Noting the “appetite in the public sector to embed these innovations,” Carlos described the framework’s value in helping organizations pursue development goals. The report underscored, he argued, why organizations “need to move beyond pilots and scale up solutions.”
Marc, meanwhile, spoke about the growing relevance of the intelligences and technologies described in the framework.
“The acceleration of technology is also accelerating the need to look into these issues,” he noted. “We’ve really been asked […] to look at digital transformation inside out and this means [avoiding] bias. [It means] making sure that there’s no blind spots when doing concept paper, in planning for a loan for governments in the region. It’s about [ensuring] those technologies really further empower ADB staff to be those available to governments in the region.”
Finally, Shachee Doshi emphasized the need to focus time and energy on machine learning. In her view, each part of the intelligence paradigm had some connection to artificial intelligence and machine learning.
“We need to recognize that AI and machine learning needs to be given a little more weight. They are all interlinked […] but [other intelligences] feed into AI and machine learning.”
Operationalizing the Four Intelligences
From this basis, the speakers shared recommendations to deepen this framework based on how they would deploy these intelligences in the regions they work.
Engaging local actors
Nii Quaynor discussed the importance of leaving sufficient time to understand local concerns and mitigate potential risks, emphasizing the importance of community engagement.
“It’s not one-size-fits-all or one side decides,” Nii said. “We are going to have to engage in more community discussions […] because many of these issues are interdisciplinary. We need to discuss [risks] and make them somehow public to develop some methods of addressing them.”
Ensuring controlled adoption and bridging the intelligence divide
Nii also spoke about the need to build up the skills of the users in development contexts, ensuring they understand and contribute to data collection, and the development organization, ensuring they understand the contexts in which they operate. Failing to promote this understanding can lead to an intelligence divide, where the capacity of technology outpaces the ability of others to mitigate the risks associated with it.
Investing privacy-preserving, representative systems
Peter Addo reiterated the importance of AI, but stressed the need for privacy-forward thinking while developing solutions. “Privacy-preserving AI solutions will be key to finding new ways of applying AI. It’s one of the next generation,” he said.
Both Shachee Doshi and Carlos Santiosio echoed the importance of this technology, but stressed the importance of good data. Shachee noted, “If we want the machine to represent the world, we need to represent the world in the data […] If we feed biases into the system, then the system is going to replicate those.”
She also discussed the need to include diverse voices into the design of AI and other systems; people should be included in conversations about their data to ensure representative work.
“Technology is moving very quickly and regulations are obviously lagging behind that. This is happening all over the world, but especially in developing countries. This leaves people behind. It leaves companies behind.”
Adopting a portfolio approach
Marc Lepage concluded the conversation by suggesting a portfolio approach to the emerging uses of technology in development, pursuing many different projects that align with general interests and shifting resources based on results.
“Once in a while, you may need to kill your darlings in terms of sunk cost. You just need to move on and find those [projects]. It doesn’t necessarily need to be about immediate benefits so much as seeing on the horizon how you can move the needle.”
While it can be tempting to create quick, agile solutions to intelligence development, there are several underlying progressions in technology that should be deeply considered before doing so.
“The whale might not be the more scaled up version of the fish,” Marc said.
Seeking Your Input
This panel provided an opportunity for The GovLab and AFD to improve its draft policy paper “Emerging Uses of Technology for Development: A New Intelligence Paradigm” ahead of final revisions November 25th and formal publication in December 2020. The GovLab invites anyone interested in this conversation or the larger research product to let us know their thoughts on the document through our peer review process or by contacting us directly at firstname.lastname@example.org.
We look forward to your input.