Are you an optimist or a pessimist when it comes to Artificial Intelligence (AI)? Do you think it will bring apocalyptic scenarios where the disruptive forces of technology take over the world and make us humans redundant or the servants of technology? Or do you think it presents opportunities for solving the world’s most pressing challenges?

Pessimist: “Extreme poverty won’t be solved by algorithms.”

Optimist:  “This Twitter sentiment dashboard will change everything.”

We at the Qatar Computing Research Institute (QCRI) and UNDP land somewhere in the middle of this spectrum. Bearing in mind the important risks of privacy breaches, ethical missteps, increased inequalities, to name a few, we see AI as an important catalyst for transforming the way we think about development and for achieving sustainable development in the 21^st century.

The question of how to harness the power of AI for social good was what we had in mind when we gathered experts from around the world for a two-day workshop in Doha, Qatar, 17-18 February 2019. Together with the private sector, academia, foundations, and international agencies, we set out to bridge the gap between the optimists and the pessimists and to address the real world challenges that can be tackled using AI. Our goal was also to build new connections to facilitate information flow across sectors and create lasting collaborations.

The workshop consisted of two components: 1) presentations that demonstrated the latest AI  for social good research. For example, how UN agencies and their national partners use data generated from social media platforms for poverty mapping and how international NGOs use satellite data to understand the characteristics and needs of displaced people. Links to presentations can be found here. 2) a hands-on session where those with development and humanitarian challenges met with AI researchers, data owners, and other AI experts to discuss how AI can be used to provide insights into their challenges.    

With the input of the AI experts, those with data-centric research challenges were able to identify new types and sources of data, new analytical approaches, and new partnerships to the problems they presented. Importantly, they also considered the risks associated with using AI, taking into account ethical, technical, political and operational issues.

We thank those holding the research challenges for presenting them, and the AI community for their interest and generosity in sharing their expertise. We welcome the AI community to peruse the research challenges; a detailed version of the list can be found here. If you would like to learn more and get involved, please contact Ingmar Weber (iweber@hbku.edu.qa) and Jennifer Colville (jennifer.colville@undp.org). 

Dr. Ingmar Weber, Research Director, Social Computing, Qatar Computing Research Institute, co-wrote this blog.