Frontier Tech for All: Testing Out a Framework for Global, Inclusive Tech Conversations

September 30, 2024

Guest blog by Gloria Gerhardt, Medhavi Hassija, Sharif Kazemi, Amy Utomo from Columbia University’s School of International and Public Affairs (SIPA), with support from Jeremy Boy, UNDP Accelerator Labs Lead Data Scientist.

Frontier technologies, such as Artificial Intelligence (AI), blockchain, the metaverse, 3D printing or brain-computer-interfaces, are all the rage. While the promises are great, there is also a global inequity in the design and implementation of these technologies, as well as in the distribution of their risks and benefits.

A key question is: who gets a say in the direction AI and other frontier technologies go? How can we support the inclusion of more diverse voices, specifically from the Global Majority, in the development and adoption of these technologies, or in shaping how they are governed, as recommended by the UN Secretary General’s AI Advisory Board?

Broadening the conversation

Discussions on frontier technologies tend to be abstract. For example, questions like “how can AI be deployed in an ethical manner?” or “how is AI going to change the nature of work?” beg specification of what is meant by AI and who is likely to be impacted. To make things more complicated, advanced technologies can sometimes go unseen. They can be embedded in complex systems, and their implications and importance can go unrecognized, even by those who use these systems. The infrastructure they require and the means through which they are deployed can also vary across locations and sectors. How can we overcome these challenges to be more inclusive in how we think about frontier technologies, and not leave our collective technological futures to those few who have access to boardrooms in Silicon Valley?

As a humble contribution to this immense task, the UNDP Accelerator Labs worked with Columbia University’s School of International and Public Affairs (SIPA) to create a Framework for Conducting National Dialogues on Frontier Technologies. This research was led by a team of SIPA graduate students, and a first prototype of the Framework, focused specifically on AI, was piloted with the UNDP Accelerator Labs in Morocco and Paraguay.

The core idea is that all voices are equal, wherever they come from, and wherever they stand in terms of technological expertise. The voice of a layperson from the Global Majority should be just as important as that of an internationally renowned expert from the Global North.

Drawing inspiration from the “Post-2015 Development Agenda: Guidelines for National Consultations,” which helped channel citizen input into what became the Sustainable Development Goals framework, we designed a framework to support public sector organizations to gather voices and perspectives on the development and deployment of frontier technologies in places that are often overlooked in global settings.

Testing it out in Morocco and Paraguay

We tested out a framework for consultations in two countries to begin to understand how AI is perceived by people of the Global Majority. We worked with UNDP Morocco as it is emerging as a significant player in the field of AI in Africa. The country is recognized for its assets in research and training and for its growing AI market. We also chose to run a test consultation in Paraguay as the country is making significant progress in promoting innovation and digital transformation in various sectors, including agriculture, transportation and healthcare.

We conducted consultations with women and men from diverse socio-economic and cultural backgrounds with varying degrees of technical knowledge:

In Morocco we spoke to 119 individuals, including members of grassroots organizations, marginalized youth, students, activists, unions, NGOs, and civil servants in three languages (Arabic, French and English)
In Paraguay, we engaged with 88 individuals, including civil servants, actors of the private sector, academics, AI experts, and farmers in rural settings mostly in Spanish, although knowledge of Guarani was needed to interpret some of the answers.

Despite the great diversity of participants, we were able to identify some common threads. The level of education of the people we consulted correlates with their overall sentiments towards AI: the higher the level of education, the more positive and the less uncertain participants were regarding the impacts of AI on their future:

When asked about whose responsibility it is to mitigate the harms of AI, people we talked to generally considered it was up to individual users and the public sector rather than, for example, the tech industry, although the reasons given for this varied a lot.
Men tended to be slightly more positive and less uncertain about AI than women.

Conducting these consultations helped us test different methodologies to refine the Framework. For more details on these pilots, as well as on the Framework’s development, please read the Background Report.

How do we structure inclusive consultations on AI?

The Consultation Framework proposes a way to anchor conversations around technological examples that are relatable to avoid getting lost in abstraction. We used a selection of consultation modalities, like in-person focus groups, street consultations, or online surveys, as well as recommendations on when and how to use them; and a series of steps for engaging representative samples of stakeholders.

graphic of consultation logic — A key part of the Framework is to level the conversation. We suggest ways to anchor consultations around frontier technology that are relatable to people.

We like to think of the Consultation Framework as a flexible foundation for structuring consultations, rather than a prescriptive checklist of action items. The recommendations provided should be adapted for specific frontier technologies (like AI or bioengineering), national contexts, or technological maturity, to ensure situated, inclusive and insightful conversations. We hope that supporting these kinds of consultations will create space for new dialogues and ultimately diversify inputs to global conversations on these technologies.

This work aims to complement UNDP’s AI Readiness Assessment which is intended for governments to take stock of the AI landscape in their country, and to assess their level of expertise across sectors. The Framework for Conducting National Dialogues on Frontier Technologies is focused on citizens and can help gather different perspectives towards the development and deployment of frontier technology, complementing the Readiness Assessment from the bottom up.

Given the growing influence of frontier technologies on everyday life and the rapid pace at which they evolve, there is an urgent need to include the input of women and men from around the world before the gap grows wider between stakeholders in the Global North and the Global Majority. We hope this draft framework is a contribution towards that end.

We invite you to read the Background Report, use the Framework and let us know about your experience to improve it with iteration.

Contributing authors to the Framework and Background Report: Ahmed Al Qutaini, Akmaral Bekbossynova, Josephine Hebling, Zhengzhou (Steve) Li, Liv Newkirk.