J-PAL’s new research and policy plan for evaluating and measuring AI implementation to fight poverty | MIT News

The Abdul Latif Jameel Poverty Action Lab (J-PAL) at MIT has funded eight new research studies to understand how artificial intelligence can be used to fight poverty through its new Project AI Evidence.

The age of AI has brought widespread optimism and skepticism about its effects on society. To realize the full potential of AI, Project Evidence of AI (PAIE) will identify which AI solutions work and for whom, and measure only the most effective, inclusive, and responsible solutions – while reducing those that may cause harm.

PAIE will generate evidence of what works by connecting governments, technology companies, and non-profit organizations with world-class economists at MIT and across J-PAL’s global networks to test and develop AI solutions to focused societal challenges.

The new initiative raises questions that policymakers are already asking: Are AI-powered teaching tools helping all children learn? How can early warning flood systems help people affected by natural disasters? Can machine learning algorithms help reduce deforestation in the Amazon? Can AI-powered chatbots help improve people’s lives? In the coming years, PAIE will run a series of funding competitions to invite proposals for testing AI tools that answer questions like these, and many others.

PAIE is financially supported by a grant from Google.org, a philanthropic grant from Community Jameel, a grant from the Canada International Development Research Center and UK International Development, and a partnership agreement with Amazon Web Services. With the help of Eric and Wendy Schmidt, given the recommendations of Schmidt Sciences, the program will also study productive AI in the workplace, especially in low and middle income countries.

Alex Diaz, head of social AI at Google.org, says, “we are happy to work with MIT and J-PAL, who are already leaders in this space, on Project AI Evidence. AI has great potential to benefit all people, but we urgently need to study what works, what doesn’t work, and why, if we want to realize this possibility.”

“Artificial intelligence has incredible potential, but only if the tools, knowledge, and capacity to shape it are accessible to all – including context-based research and evidence of what works and doesn’t work,” added Maggie Gorman-Velez, vice president of strategy, regions, and policies at IDRC. “That’s why IDRC is proud to support this new pilot project as part of our ongoing commitment to the responsible measurement of proven AI innovations that are safe, inclusive, and locally appropriate.”

J-PAL is uniquely positioned to help understand the effects of AI on society: Since its founding in 2003, the J-PAL network of researchers has led more than 2,500 rigorous evaluations of public policies and programs around the world. Through PAIE, J-PAL will bring together leading experts in AI technology, research, and social policy, in line with MIT president Sally Kornbluth’s focus on productive AI as a priority.

PAIE is led by Professor Joshua Blumenstock of the University of California at Berkeley; J-PAL Global Managing Director Iqbal Dhaliwal; and Professor David Yangangwa-Drott from the University of Zurich.

A new examination of urgent policy questions

The studies funded in the first round of the PAIE competition examine urgent questions in key areas such as education, health, climate, and economic opportunity.

How can AI work more effectively in classrooms, helping students and teachers?

Existing research shows that personalized learning is important for students, but it is challenging to implement it with limited resources. In Kenya, social education business EIDU has developed an AI tool that helps teachers identify learning gaps and adjust their daily lesson plans. In India, the non-governmental organization (NGO) Pratham is developing an AI tool to increase the impact and scale of Evidence-informed Teaching in a Right-of-Way approach. J-PAL researchers Daron Acemoglu, Iqbal Dhaliwal, and Francisco Gallego will work with both organizations to study the effects and potential of these different use cases on teacher productivity and student learning.

Can AI tools reduce gender bias in schools?

Researchers are working with the Italian Ministry of Education to test whether AI tools can help close the gender gap in student performance by addressing unconscious teacher bias. J-PAL Undergraduates Michela Carlana and Will Dobbie, along with Francesca Miserocchi and Eleonora Patacchini, will research the effects of two AI tools, one that helps teachers predict performance and the second that provides real-time feedback on the variability of their decisions.

Can AI help career counselors identify career opportunities?

In Kenya, researchers are testing whether an AI tool can identify neglected skills and open up employment opportunities, especially for youth, women, and those without formal education. Working with NGOs Swahilipot and Tabiya, Jasmine Baier and J-PAL researcher Christian Meyer will explore how this tool changes people’s job search and employment strategies. This research will illuminate AI as a complement to, rather than a replacement for, human technology in task management.

I’m looking forward

As the use of AI in the social sector grows, this assessment is the first step towards finding effective and responsible solutions that will lead to the eradication of poverty and inequality.

J-PAL’s Dhaliwal comments, “IJ-PAL has a long history of exploring new technologies and their ability to improve people’s lives. While AI has great potential, we need to maximize its benefits and minimize the potential harms of innovation.”

IJ-PAL is also looking for new collaborators who share its vision of discovering and scaling real-world AI solutions. It aims to support more governments and social sector organizations that want to use AI responsibly, and will continue to increase funding for new research and provide policy guidance based on the latest research.

To learn more about the Evidence for AI Project, sign up for the J-PAL newsletter or contact paie@povertyactionlab.org.