This story appears as a part of Open Data PGH, a joint reporting project by Technical.ly and PublicSource on open data trends in Pittsburgh, underwritten by Heinz Endowments. Learn more here and get updates here.
On May 15 through 17, a diverse group of policymakers, artificial-intelligence practitioners, researchers, industry representatives and domain experts from over 60 countries gathered in Geneva, Switzerland for the second annual AI for Good Global Summit.
The goal was to formulate practical strategies and projects that ensure trusted, safe and inclusive development of AI technologies and equitable access to their benefits for all people and communities around the world.
(I’ve been the Pittsburgh liaison for AI for Good initiatives since attending the inaugural AI for Good Summit in 2017, and led the Pittsburgh delegation for the 2018 summit.)
Since the inaugural summit in 2017, the AI for Good series has been the leading United Nations platform for dialogue on AI, focusing on impactful AI solutions with long-term benefits that align with the U.N.’s 17 sustainable development goals. Organized by the International Telecommunication Union (ITU) in partnership with XPRIZE and the Association for Computing Machinery (ACM), this year’s summit also brought together 32 other U.N. agencies, including the World Health Organization, World Bank, UNICEF, and the U.N. Office for Disarmament Affairs.
Over the course of three days, summit participants engaged in working sessions focused on developing projects in four primary topic tracks: AI and Satellite Imagery, AI and Health, AI and Smart Cities and Trust in AI. In total, leaders from across these tracks proposed 35 distinct projects. One additional track called AI and Data Commons deployed a team of “data rapporteurs” (which included myself) to collect insights from each of the other four tracks and synthesize a roadmap for the development of “data commons.” As a means of facilitating the collection, cleaning, labeling and use of AI tools and datasets, the data commons represents an exciting prospect for any civic coder or supporter of open data initiatives.
Of the projects proposed, most aimed to address specific issue areas by leveraging existing research, data and initiatives. For example, one project focused on using satellite data to support micro-insurance programs for small-hold farmers in Africa. A few of the health projects explored building on current AI research to help detect and evaluate conditions such as vision loss, osteoarthritis and malnutrition, especially in developing countries.
Other projects were more foundational, focused on creating platforms to facilitate communication and collaboration among experts and stakeholders worldwide. An “Internet of Cities” was proposed, as a network where data, knowledge, and expertise from both successful and failed Smart City practices could be shared to inform future initiatives. Another project, called TrustFactory.ai, was launched at the summit to serve as a convener of cross-cultural dialogue around trust in AI, with the aim of informing inclusive global standards.
Underlying all of these projects was a clear need for not just a vast quantity of data, but quality data that is accurate, robust, diverse and trustworthy. The data rapporteurs, led by Urs Gasser from Harvard University’s Berkman Klein Center, developed a framework that shows how a well-designed data commons should be more than a just a data repository, and should account for details relating to the organizations, laws and human beings that might affect affect or be affected by each dataset.
Furthermore, Amir Banifatemi, New Frontiers lead at XPRIZE and co-organizer of the AI for Good Summit, shared plans for the development of an AI Commons organization. The AI Commons aims to support and enhance AI for Good projects by enabling people from around the world to access data, discuss ideas, propose problems and contribute solutions. As a first step, the AI Commons is building a data commons, using a technical infrastructure designed by the company Ocean Protocol, which specializes in decentralized data pooling and data exchange.
The projects outlined at the AI for Good Summit are all at nascent stages of development, and their success will require the support and leadership of individuals, organizations, cities, and other stakeholder groups around world. Pittsburgh has already stepped up and committed to fulfill the responsibility of an AI for Good city.
At the AI for Good Summit, Mayor William Peduto was invited to call the opening of the “Beneficial AI Era” on behalf of Pittsburgh. The city was recognized for its efforts to balance the development of cutting-edge technologies with priorities around ethical and inclusive innovation. Mayor Peduto also cited lessons from the city’s industrial past and specific examples from his current OnePGH resilience strategy to reinforce how well-poised Pittsburgh is to lead the AI for Good Movement.
What’s become imminently clear is that this movement will be shaped from the ground up, led by the cities, researchers and civic technologists who have the desire and ability to use these technologies to solve real problems. Among those currently leading the way in Pittsburgh are David Danks from Carnegie Mellon University, Diane Litman from the University of Pittsburgh and Shinjini Kundu from the National Institutes of Health, who also attended the summit.
We’ll need many more to join the ranks if we’re to fully capture the opportunities and mitigate the risks that artificial intelligence presents to our society.-30-
Big tech companies are flocking to Pittsburgh. The foundation was laid over decades
An all-staff letter from Technically Media’s CEO about our values
3 technologists’ hopes for a more open Pittsburgh
How to build a civic tech community, according to Pittsburgh leaders
Sign-up for daily news updates from Technical.ly Pittsburgh